cellular-infrastructure
/
osmo-bsc
mirror of https://gerrit.osmocom.org/osmo-bsc
9
0
Fork 0
Code Issues Releases Wiki Activity
osmo-bsc/src/osmo-bsc/assignment_fsm.c

1011 lines
35 KiB
C
Raw Normal View History

large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
/* osmo-bsc BSSMAP Assignment procedure implementation.
*
* (C) 2018 by sysmocom - s.f.m.c. GmbH <info@sysmocom.de>
* All Rights Reserved
*
* Author: Neels Hofmeyr <neels@hofmeyr.de>
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU Affero General Public License as published by
* the Free Software Foundation; either version 3 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU Affero General Public License for more details.
*
* You should have received a copy of the GNU Affero General Public License
* along with this program. If not, see <http://www.gnu.org/licenses/>.
*
*/
use libosmocore osmo_tdef Move the T_defs API to libosmocore as osmo_tdefs: remove the local T_defs API and use libosmocore's osmo_tdef* API instead. The root reason is moving the mgw_endpoint_fsm to libosmo-mgcp-client to be able to use it in osmo-msc for inter-MSC handover. When adding osmo_tdef, the new concept of timer groups was added to the API. It would make sense to apply group names here as well, but do not modify the VTY configuration for timers. The future might bring separate groups (or not). Depends: Ibd6b1ed7f1bd6e1f2e0fde53352055a4468f23e5 (libosmocore) Change-Id: I66674a5d8403d820038762888c846bae10ceac58
2019-01-28 02:52:14 +00:00
#include <osmocom/core/tdef.h>
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
#include <osmocom/gsm/gsm0808.h>
move mgw endpoint FSM to osmo-mgw.git osmo-mgw.git also includes fixes of the MGW endpoint FSM, for example I92a9944acc96398acd6649f9c3c5badec5dd6dcc. Depends: I9a3effd38e72841529df6c135c077116981dea36 (osmo-mgw) Change-Id: I03e6b48d9b0a5370310d5f56809259ff7909cf9d
2019-04-23 16:37:37 +00:00
#include <osmocom/mgcp_client/mgcp_client_endpoint_fsm.h>
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
#include <osmocom/bsc/debug.h>
#include <osmocom/bsc/gsm_data.h>
#include <osmocom/bsc/lchan_fsm.h>
#include <osmocom/bsc/bsc_subscr_conn_fsm.h>
#include <osmocom/bsc/osmo_bsc_lcls.h>
#include <osmocom/bsc/bsc_msc_data.h>
cosmetic: rename bsc_api.h to gsm_08_08.h See also I91922f557072d0fb8cfe213a8a7b50f3bb23dea0, which renames osmo_bsc_api.c to gsm_08_08.c. Change-Id: I7179eb27183ee213f8fc8d548895b67aa43dc6a2
2018-07-24 16:10:05 +00:00
#include <osmocom/bsc/gsm_08_08.h>
BS-11: Send proprietary MRPCI message after assignment + HO complete This is a non-standard RSL message that must be sent after each traffic channel has been established. Without it, any voice call will be disconnected within seconds. This is a hack; we need to store the subscribers classmark2 value and use it here. Change-Id: I6cb6d25e405aa888c4df4022d897330a6af9e946 Related: OS#2389
2020-08-01 15:03:31 +00:00
#include <osmocom/bsc/gsm_04_08_rr.h>
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
#include <osmocom/bsc/lchan_select.h>
#include <osmocom/bsc/abis_rsl.h>
Move struct gsm_bts: gsm_data.* => bts.* Place all code related to the object into the related file. Having all the data model in one file made sense in early stage of development to make progress quickly, but nowadays it hurts more than helps, due to constantly growing size and more and more bits being added to the model, gaining in complexity. Currently, having lots of different objects mixed up in gsm_data.h is a hole of despair, where nobody can make any sense were to properly put new stuff in, ending up with functions related to same object in different files or with wrong prefixes, declarations of non-existing functions, etc. because people cannot make up their mind on strict relation to objects in the data model. Splitting them in files really helps finding code operating on a specific object and helping with logically splitting in the future. Change-Id: I00c15f5285b5c1a0109279b7ab192d5467a04ece
2020-07-15 18:53:16 +00:00
#include <osmocom/bsc/bts.h>
move BSC level stats and rate counters to new bsc_stats.[hc] Instead of having static const structs in header files (which end up duplicated in each and every compile unit!), have one .c file with the rate_ctr and stat_item descriptions. Related: SYS#5542 Change-Id: I8fd6380b5ae8ed2d3347e7cfbf674c30b6841ed9
2021-08-29 11:18:40 +00:00
#include <osmocom/bsc/bsc_stats.h>
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
#include <osmocom/bsc/assignment_fsm.h>
static struct osmo_fsm assignment_fsm;
struct gsm_subscriber_connection *assignment_fi_conn(struct osmo_fsm_inst *fi)
{
OSMO_ASSERT(fi);
OSMO_ASSERT(fi->fsm == &assignment_fsm);
OSMO_ASSERT(fi->priv);
return fi->priv;
}
use libosmocore osmo_tdef Move the T_defs API to libosmocore as osmo_tdefs: remove the local T_defs API and use libosmocore's osmo_tdef* API instead. The root reason is moving the mgw_endpoint_fsm to libosmo-mgcp-client to be able to use it in osmo-msc for inter-MSC handover. When adding osmo_tdef, the new concept of timer groups was added to the API. It would make sense to apply group names here as well, but do not modify the VTY configuration for timers. The future might bring separate groups (or not). Depends: Ibd6b1ed7f1bd6e1f2e0fde53352055a4468f23e5 (libosmocore) Change-Id: I66674a5d8403d820038762888c846bae10ceac58
2019-01-28 02:52:14 +00:00
static const struct osmo_tdef_state_timeout assignment_fsm_timeouts[32] = {
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
[ASSIGNMENT_ST_WAIT_LCHAN_ACTIVE] = { .T=10 },
[ASSIGNMENT_ST_WAIT_RR_ASS_COMPLETE] = { .keep_timer=true },
[ASSIGNMENT_ST_WAIT_LCHAN_ESTABLISHED] = { .keep_timer=true },
[ASSIGNMENT_ST_WAIT_MGW_ENDPOINT_TO_MSC] = { .T=23042 },
};
/* Transition to a state, using the T timer defined in assignment_fsm_timeouts.
* The actual timeout value is in turn obtained from network->T_defs.
* Assumes local variable fi exists. */
#define assignment_fsm_state_chg(state) \
use libosmocore osmo_tdef Move the T_defs API to libosmocore as osmo_tdefs: remove the local T_defs API and use libosmocore's osmo_tdef* API instead. The root reason is moving the mgw_endpoint_fsm to libosmo-mgcp-client to be able to use it in osmo-msc for inter-MSC handover. When adding osmo_tdef, the new concept of timer groups was added to the API. It would make sense to apply group names here as well, but do not modify the VTY configuration for timers. The future might bring separate groups (or not). Depends: Ibd6b1ed7f1bd6e1f2e0fde53352055a4468f23e5 (libosmocore) Change-Id: I66674a5d8403d820038762888c846bae10ceac58
2019-01-28 02:52:14 +00:00
osmo_tdef_fsm_inst_state_chg(fi, state, \
assignment_fsm_timeouts, \
((struct gsm_subscriber_connection*)(fi->priv))->network->T_defs, \
5)
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
/* Log failure and transition to ASSIGNMENT_ST_FAILURE, which triggers the appropriate actions. */
#define assignment_fail(cause, fmt, args...) do { \
struct gsm_subscriber_connection *_conn = fi->priv; \
_conn->assignment.failure_cause = cause; \
LOG_ASSIGNMENT(_conn, LOGL_ERROR, "Assignment failed in state %s, cause %s: " fmt "\n", \
osmo_fsm_inst_state_name(fi), gsm0808_cause_name(cause), ## args); \
Count assignment rates per BTS as well This adds the assignment counters for the BTS as well and changes the assignment_count() macro to increase both the counters for the BSC as well as the BTS. Related: SYS#4877 Change-Id: I0009e51d4caf68e762138d98e2e23d49acc3cc1a
2020-06-08 12:36:46 +00:00
assignment_count_result(CTR_ASSIGNMENT_ERROR); \
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
on_assignment_failure(_conn); \
for linter: s/while(0)/while (0) Change-Id: Ib422e7d1a7d543dcd8738581839ce55bb8fc29d2
2021-11-06 16:01:58 +00:00
} while (0)
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
/* Assume presence of local var 'conn' as struct gsm_subscriber_connection */
#define assignment_count(counter) do { \
Count assignment rates per BTS as well This adds the assignment counters for the BTS as well and changes the assignment_count() macro to increase both the counters for the BSC as well as the BTS. Related: SYS#4877 Change-Id: I0009e51d4caf68e762138d98e2e23d49acc3cc1a
2020-06-08 12:36:46 +00:00
struct gsm_bts *bts = conn_get_bts(conn); \
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
LOG_ASSIGNMENT(conn, LOGL_DEBUG, "incrementing rate counter: %s %s\n", \
Count assignment rates per BTS as well This adds the assignment counters for the BTS as well and changes the assignment_count() macro to increase both the counters for the BSC as well as the BTS. Related: SYS#4877 Change-Id: I0009e51d4caf68e762138d98e2e23d49acc3cc1a
2020-06-08 12:36:46 +00:00
bsc_ctr_description[BSC_##counter].name, \
bsc_ctr_description[BSC_##counter].description); \
Use new stat item/ctr getter APIs Generated with following and similar spatch snippets: """ @@ expression E1, E2; @@ - &E2->ctr[E1] + rate_ctr_group_get_ctr(E2, E1) """ Change-Id: I0b43f922a595d694ac0aeda80107ef9bf4e755e7
2021-06-04 15:30:21 +00:00
rate_ctr_inc(rate_ctr_group_get_ctr(conn->network->bsc_ctrs, BSC_##counter)); \
stats: add SIGN/SPEECH assignment subcategories Change-Id: I73f4dab6edb0951180f2bbcfc352ff34de647679
2021-02-17 12:04:29 +00:00
if (bts) { \
Use new stat item/ctr getter APIs Generated with following and similar spatch snippets: """ @@ expression E1, E2; @@ - &E2->ctr[E1] + rate_ctr_group_get_ctr(E2, E1) """ Change-Id: I0b43f922a595d694ac0aeda80107ef9bf4e755e7
2021-06-04 15:30:21 +00:00
rate_ctr_inc(rate_ctr_group_get_ctr(bts->bts_ctrs, BTS_##counter)); \
ensure chan_mode comparisons in non-VAMOS mode Both VAMOS- and non-VAMOS speech modes should result in indentical voice handling. So make sure that all chan_modes are converted to non-vamos before comparing / evaluating in switch statements. Change-Id: I791e7966b1f8eaa3299a8a46abeb313cf5136e0b
2021-05-28 17:18:11 +00:00
switch (gsm48_chan_mode_to_non_vamos(conn->assignment.req.ch_mode_rate_list[0].chan_mode)) { \
remove special case from assignment_count_result() The single place that calls this special case is now actually doing the exact same thing as the general case, and this is merely code dup. Related: SYS#5315 OS#4940 Change-Id: Ic62e8a48d3c88a1966086240a41732d169328491
2021-04-28 15:21:04 +00:00
case GSM48_CMODE_SIGN: \
Use new stat item/ctr getter APIs Generated with following and similar spatch snippets: """ @@ expression E1, E2; @@ - &E2->ctr[E1] + rate_ctr_group_get_ctr(E2, E1) """ Change-Id: I0b43f922a595d694ac0aeda80107ef9bf4e755e7
2021-06-04 15:30:21 +00:00
rate_ctr_inc(rate_ctr_group_get_ctr(bts->bts_ctrs, BTS_##counter##_SIGN)); \
dbg log: also log assignment counters on BTS level Change-Id: I68b3e794ddf97c80f647c3536ec466217b338af9
2021-11-29 15:24:55 +00:00
LOG_ASSIGNMENT(conn, LOGL_DEBUG, "incrementing rate counter: bts%u %s %s\n", \
bts->nr, \
bts_ctr_description[BTS_##counter##_SIGN].name, \
bts_ctr_description[BTS_##counter##_SIGN].description); \
remove special case from assignment_count_result() The single place that calls this special case is now actually doing the exact same thing as the general case, and this is merely code dup. Related: SYS#5315 OS#4940 Change-Id: Ic62e8a48d3c88a1966086240a41732d169328491
2021-04-28 15:21:04 +00:00
break; \
case GSM48_CMODE_SPEECH_V1: \
case GSM48_CMODE_SPEECH_EFR: \
case GSM48_CMODE_SPEECH_AMR: \
Use new stat item/ctr getter APIs Generated with following and similar spatch snippets: """ @@ expression E1, E2; @@ - &E2->ctr[E1] + rate_ctr_group_get_ctr(E2, E1) """ Change-Id: I0b43f922a595d694ac0aeda80107ef9bf4e755e7
2021-06-04 15:30:21 +00:00
rate_ctr_inc(rate_ctr_group_get_ctr(bts->bts_ctrs, BTS_##counter##_SPEECH)); \
dbg log: also log assignment counters on BTS level Change-Id: I68b3e794ddf97c80f647c3536ec466217b338af9
2021-11-29 15:24:55 +00:00
LOG_ASSIGNMENT(conn, LOGL_DEBUG, "incrementing rate counter: bts%u %s %s\n", \
bts->nr, \
bts_ctr_description[BTS_##counter##_SPEECH].name, \
bts_ctr_description[BTS_##counter##_SPEECH].description); \
remove special case from assignment_count_result() The single place that calls this special case is now actually doing the exact same thing as the general case, and this is merely code dup. Related: SYS#5315 OS#4940 Change-Id: Ic62e8a48d3c88a1966086240a41732d169328491
2021-04-28 15:21:04 +00:00
break; \
default: \
break; \
stats: add SIGN/SPEECH assignment subcategories Change-Id: I73f4dab6edb0951180f2bbcfc352ff34de647679
2021-02-17 12:04:29 +00:00
} \
} \
for linter: s/while(0)/while (0) Change-Id: Ib422e7d1a7d543dcd8738581839ce55bb8fc29d2
2021-11-06 16:01:58 +00:00
} while (0)
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
#define assignment_count_result(counter) do { \
if (!conn->assignment.result_rate_ctr_done) { \
assignment_count(counter); \
conn->assignment.result_rate_ctr_done = true; \
} else \
LOG_ASSIGNMENT(conn, LOGL_DEBUG, \
fix another log line end in assignment_fsm.c Change-Id: I3be062ad7ecb5ba801cd7412e90c4bc5bf7e367c
2019-03-15 20:06:25 +00:00
"result rate counter already recorded, NOT counting as: %s %s\n", \
Count assignment rates per BTS as well This adds the assignment counters for the BTS as well and changes the assignment_count() macro to increase both the counters for the BSC as well as the BTS. Related: SYS#4877 Change-Id: I0009e51d4caf68e762138d98e2e23d49acc3cc1a
2020-06-08 12:36:46 +00:00
bsc_ctr_description[BSC_##counter].name, \
bsc_ctr_description[BSC_##counter].description); \
for linter: s/while(0)/while (0) Change-Id: Ib422e7d1a7d543dcd8738581839ce55bb8fc29d2
2021-11-06 16:01:58 +00:00
} while (0)
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
void assignment_reset(struct gsm_subscriber_connection *conn)
{
if (conn->assignment.new_lchan) {
struct gsm_lchan *lchan = conn->assignment.new_lchan;
conn->assignment.new_lchan = NULL;
Send EUTRAN neighs based on whether Common Id msg contained Last used E-UTRAN PLMN ID From 3GPP TS 48.008 sec 3.1.30 "Common ID": """ If the SCCP connection is established due to CSFB from E-UTRAN and the MSC supports return to the last used PLMN after CS fallback, then it should send the COMMON ID message to the BSS including the Last used E-UTRAN PLMN ID information element if available at the MSC immediately following the successful SCCP connection setup. """ Furthermore, 3GPP TS 48.008 version 16.0.0 Release 16 "3.2.1.21 CLEAR COMMAND", for field CSFB Indication, states: """ NOTE: This information element doesn't serve any useful purpose. MSCs should not send the information element unless it is required by the recipients (due to the need to interwork with older versions of the protocol). It is expected that in future versions of the present document, this information element will be deleted from this message. """ Hence, build up the EUTRAN neighbor list based on whether we received the Last E-UTRAN PLMN ID IE during Common Id. In the future, we should probably filter the list while populating it based on the received IE. This change will also allow reusing same mechanism for SRVCC EUTRAN->GERAN support, where te Last E-UTRAN PLMN ID IE can be found inside "Old BSS to New BSS information" IE in msg HANDOVER REQUEST. Related: SYS#5337 Change-Id: I5d290ac55eca5adde1c33396422f4c10b83c03d5
2021-04-12 18:19:30 +00:00
lchan_release(lchan, false, true, RSL_ERR_EQUIPMENT_FAIL, NULL);
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
}
if (conn->assignment.created_ci_for_msc) {
fix rare segfault in MGCP client handling Add missing conn->assignment.created_ci_for_msc to gscon_forget_mgw_endpoint_ci(). Before this patch, when assignment.created_ci_for_msc lingers after a DLCX, it can cause a use-after-free on assignment_reset(). Possible scenario is rx BSSMAP Clear Cmd during ongoing Assignment. In assignment_reset(), locally cache the ci pointer, because gscon_forget_mgw_endpoint_ci() now NULLs created_ci_for_msc. Related: OS#5572 Change-Id: If89610020f47fd6517081dd11b83911b043bd0f1
2022-05-30 22:16:39 +00:00
/* Store ci pointer locally, because gscon_forget_mgw_endpoint_ci() NULLs
* conn->assignment.created_ci_for_msc. */
struct osmo_mgcpc_ep_ci *ci = conn->assignment.created_ci_for_msc;
gscon_forget_mgw_endpoint_ci(conn, ci);
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
/* If this is the last endpoint released, the mgw_endpoint_fsm will terminate and tell
* the gscon about it. */
fix rare segfault in MGCP client handling Add missing conn->assignment.created_ci_for_msc to gscon_forget_mgw_endpoint_ci(). Before this patch, when assignment.created_ci_for_msc lingers after a DLCX, it can cause a use-after-free on assignment_reset(). Possible scenario is rx BSSMAP Clear Cmd during ongoing Assignment. In assignment_reset(), locally cache the ci pointer, because gscon_forget_mgw_endpoint_ci() now NULLs created_ci_for_msc. Related: OS#5572 Change-Id: If89610020f47fd6517081dd11b83911b043bd0f1
2022-05-30 22:16:39 +00:00
osmo_mgcpc_ep_ci_dlcx(ci);
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
}
conn->assignment = (struct assignment_fsm_data){
.fi = conn->assignment.fi, /* The FSM shall clear itself when it's done. */
};
}
static void on_assignment_failure(struct gsm_subscriber_connection *conn)
{
assignment_fsm: send BSSMAP response only after Assignment Request So far, only the MSC asked for Assignment via Assignment Request, which we answer with a BSSMAP Assignment Complete or Assignment Failure when done. When Assignment is triggered for any other reason (congestion resolution, VAMOS, VTY), we will not send any such messages to the MSC. Additional enum values will be added in subsequent commits: Id56a890106b93fcee67ac9401b890e7b63bba421 ASSIGN_FOR_CONGESTION_RESOLUTION If006f5caaf83b07675f57e5665cfa79328da55e6 ASSIGN_FOR_VTY Related: SYS#5315 OS#4940 Change-Id: Ie0cddbdb00abcec78e153f4ae6d04ce75080a111
2021-05-01 02:52:05 +00:00
/* Send Assignment Failure to MSC only when the assignment was requested via BSSAP. Do not send anything to the
* MSC if re-assignment was requested for congestion resolution, for VAMOS multiplexing, or by VTY. */
if (conn->assignment.req.assign_for == ASSIGN_FOR_BSSMAP_REQ) {
struct msgb *resp = gsm0808_create_assignment_failure(conn->assignment.failure_cause, NULL);
if (!resp) {
LOG_ASSIGNMENT(conn, LOGL_ERROR, "Unable to compose BSSMAP Assignment Failure message\n");
} else {
Use new stat item/ctr getter APIs Generated with following and similar spatch snippets: """ @@ expression E1, E2; @@ - &E2->ctr[E1] + rate_ctr_group_get_ctr(E2, E1) """ Change-Id: I0b43f922a595d694ac0aeda80107ef9bf4e755e7
2021-06-04 15:30:21 +00:00
rate_ctr_inc(rate_ctr_group_get_ctr(conn->sccp.msc->msc_ctrs, MSC_CTR_BSSMAP_TX_DT1_ASSIGMENT_FAILURE));
assignment_fsm: send BSSMAP response only after Assignment Request So far, only the MSC asked for Assignment via Assignment Request, which we answer with a BSSMAP Assignment Complete or Assignment Failure when done. When Assignment is triggered for any other reason (congestion resolution, VAMOS, VTY), we will not send any such messages to the MSC. Additional enum values will be added in subsequent commits: Id56a890106b93fcee67ac9401b890e7b63bba421 ASSIGN_FOR_CONGESTION_RESOLUTION If006f5caaf83b07675f57e5665cfa79328da55e6 ASSIGN_FOR_VTY Related: SYS#5315 OS#4940 Change-Id: Ie0cddbdb00abcec78e153f4ae6d04ce75080a111
2021-05-01 02:52:05 +00:00
gscon_sigtran_send(conn, resp);
}
stats: Add counters for Tx BSSMAP messages. We already have counters for Rx side, now we also count Tx side. See comments in the msc_ctr_description array implementation for the details. Change-Id: I89a173f6bdd9a3c21233fe01d07ab2ff0442bb10
2020-05-09 18:27:40 +00:00
}
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
/* If assignment failed as early as in assignment_fsm_start(), there may not be an fi yet. */
if (conn->assignment.fi) {
LOG_ASSIGNMENT(conn, LOGL_ERROR, "Assignment failed\n");
osmo_fsm_inst_term(conn->assignment.fi, OSMO_FSM_TERM_ERROR, 0);
}
}
bssap: Feed Assign Complete with Osmux CID retrieved from MGW The CID is sent back to the MSC so it can configure its MGW to send Osmux frames to the BSC MGW using this CID. Change-Id: I77dfdd965ae828c39a9818669177aefd22bc02f2
2019-04-16 18:33:25 +00:00
static void _gsm0808_ass_compl_extend_osmux(struct msgb *msg, uint8_t cid)
{
OSMO_ASSERT(msg->l3h[1] == msgb_l3len(msg) - 2); /*TL not in len */
msgb_tv_put(msg, GSM0808_IE_OSMO_OSMUX_CID, cid);
msg->l3h[1] = msgb_l3len(msg) - 2;
}
cosmetic: drop param from send_assignment_complete() After previous patch I8dd561d744d8081b5ac5ffa7635f17ac19bcda45, all callers pass conn->lchan to send_assignment_complete(), so there is no need to pass it separately from conn. Change-Id: I3c038d91a53c81d9fcf4ec62f853b59bb1ecd244
2018-09-24 15:38:37 +00:00
static void send_assignment_complete(struct gsm_subscriber_connection *conn)
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
{
int rc;
struct gsm0808_speech_codec sc;
struct gsm0808_speech_codec *sc_ptr = NULL;
struct sockaddr_storage addr_local;
struct sockaddr_storage *addr_local_p = NULL;
bssap: Feed Assign Complete with Osmux CID retrieved from MGW The CID is sent back to the MSC so it can configure its MGW to send Osmux frames to the BSC MGW using this CID. Change-Id: I77dfdd965ae828c39a9818669177aefd22bc02f2
2019-04-16 18:33:25 +00:00
uint8_t osmux_cid = 0;
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
int perm_spch = 0;
uint8_t chosen_channel;
struct msgb *resp;
cosmetic: drop param from send_assignment_complete() After previous patch I8dd561d744d8081b5ac5ffa7635f17ac19bcda45, all callers pass conn->lchan to send_assignment_complete(), so there is no need to pass it separately from conn. Change-Id: I3c038d91a53c81d9fcf4ec62f853b59bb1ecd244
2018-09-24 15:38:37 +00:00
struct gsm_lchan *lchan = conn->lchan;
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
struct osmo_fsm_inst *fi = conn->fi;
make sure channel mode and s15_s0 are updated only after an ACK I noticed during testing that an lchan used as TCH/F in fact still had its channel mode set to Signalling -- because on Assignment, the Speech mode used to be placed in the *previous* lchan and the new lchan was never updated after the Activ ACK. This is unbearable confusion which I complained about numerous times, so far mostly for cosmetic reasons. But implementing re-assignment properly actually requires this to be cleaned up. Keep all volatile chan mode settings in lchan->activate.* or lchan->modify.*, and only update lchan->* members when an ACK has been received for those settings. So a failed request keeps a sane state. Make sure that those settings are in fact updated in the proper lchan, upon an ACK, so that subsequent re-assignment or mode-modify know the accurate lchan state. Related are upcoming patches that sort out the AMR multirate configuration in a similar fashion, see Iebac2dc26412d877e5364f90d6f2ed7a7952351e Ia7519d2fa9e7f0b61b222d27d077bde4660c40b9 Ie57f9d0e3912632903d9740291225bfd1634ed47. Related: SYS#5315 OS#4940 OS#3787 OS#3833 Change-Id: Ie0da36124d73efc28a8809b63d7c96e2167fc412
2021-04-27 23:17:14 +00:00
chosen_channel = gsm0808_chosen_channel(lchan->type, lchan->current_ch_mode_rate.chan_mode);
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
if (!chosen_channel) {
assignment_fail(GSM0808_CAUSE_EQUIPMENT_FAILURE,
"Unable to compose Chosen Channel for mode=%s type=%s",
make sure channel mode and s15_s0 are updated only after an ACK I noticed during testing that an lchan used as TCH/F in fact still had its channel mode set to Signalling -- because on Assignment, the Speech mode used to be placed in the *previous* lchan and the new lchan was never updated after the Activ ACK. This is unbearable confusion which I complained about numerous times, so far mostly for cosmetic reasons. But implementing re-assignment properly actually requires this to be cleaned up. Keep all volatile chan mode settings in lchan->activate.* or lchan->modify.*, and only update lchan->* members when an ACK has been received for those settings. So a failed request keeps a sane state. Make sure that those settings are in fact updated in the proper lchan, upon an ACK, so that subsequent re-assignment or mode-modify know the accurate lchan state. Related are upcoming patches that sort out the AMR multirate configuration in a similar fashion, see Iebac2dc26412d877e5364f90d6f2ed7a7952351e Ia7519d2fa9e7f0b61b222d27d077bde4660c40b9 Ie57f9d0e3912632903d9740291225bfd1634ed47. Related: SYS#5315 OS#4940 OS#3787 OS#3833 Change-Id: Ie0da36124d73efc28a8809b63d7c96e2167fc412
2021-04-27 23:17:14 +00:00
get_value_string(gsm48_chan_mode_names, lchan->current_ch_mode_rate.chan_mode),
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
gsm_lchant_name(lchan->type));
return;
}
/* Generate voice related fields */
if (conn->assignment.requires_voice_stream) {
make sure channel mode and s15_s0 are updated only after an ACK I noticed during testing that an lchan used as TCH/F in fact still had its channel mode set to Signalling -- because on Assignment, the Speech mode used to be placed in the *previous* lchan and the new lchan was never updated after the Activ ACK. This is unbearable confusion which I complained about numerous times, so far mostly for cosmetic reasons. But implementing re-assignment properly actually requires this to be cleaned up. Keep all volatile chan mode settings in lchan->activate.* or lchan->modify.*, and only update lchan->* members when an ACK has been received for those settings. So a failed request keeps a sane state. Make sure that those settings are in fact updated in the proper lchan, upon an ACK, so that subsequent re-assignment or mode-modify know the accurate lchan state. Related are upcoming patches that sort out the AMR multirate configuration in a similar fashion, see Iebac2dc26412d877e5364f90d6f2ed7a7952351e Ia7519d2fa9e7f0b61b222d27d077bde4660c40b9 Ie57f9d0e3912632903d9740291225bfd1634ed47. Related: SYS#5315 OS#4940 OS#3787 OS#3833 Change-Id: Ie0da36124d73efc28a8809b63d7c96e2167fc412
2021-04-27 23:17:14 +00:00
perm_spch = gsm0808_permitted_speech(lchan->type, lchan->current_ch_mode_rate.chan_mode);
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
if (gscon_is_aoip(conn)) {
move mgw endpoint FSM to osmo-mgw.git osmo-mgw.git also includes fixes of the MGW endpoint FSM, for example I92a9944acc96398acd6649f9c3c5badec5dd6dcc. Depends: I9a3effd38e72841529df6c135c077116981dea36 (osmo-mgw) Change-Id: I03e6b48d9b0a5370310d5f56809259ff7909cf9d
2019-04-23 16:37:37 +00:00
if (!osmo_mgcpc_ep_ci_get_crcx_info_to_sockaddr(conn->user_plane.mgw_endpoint_ci_msc,
&addr_local)) {
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
assignment_fail(GSM0808_CAUSE_EQUIPMENT_FAILURE,
"Unable to compose RTP address of MGW -> MSC");
return;
}
addr_local_p = &addr_local;
}
SCCPlite Assignment Complete: include Speech Codec (Chosen) Compose the Speech Codec (Chosen) IE not only for AoIP, but also for SCCPlite: place the code that assigns the codec to sc_ptr outside of the gscon_is_aoip() if scope. This way the MSC is told the chosen speech codec, which is mandatory for IP based user plane, and was missing until now. Related: OS#3528 Change-Id: Ibedade8d71a7994d25a63bc2faa2a24a10bfffa1
2018-09-06 13:39:21 +00:00
bssap: Feed Assign Complete with Osmux CID retrieved from MGW The CID is sent back to the MSC so it can configure its MGW to send Osmux frames to the BSC MGW using this CID. Change-Id: I77dfdd965ae828c39a9818669177aefd22bc02f2
2019-04-16 18:33:25 +00:00
if (gscon_is_aoip(conn) && conn->assignment.req.use_osmux) {
if (!osmo_mgcpc_ep_ci_get_crcx_info_to_osmux_cid(conn->user_plane.mgw_endpoint_ci_msc,
&osmux_cid)) {
assignment_fail(GSM0808_CAUSE_EQUIPMENT_FAILURE,
"Unable to compose Osmux CID of MGW -> MSC");
return;
}
}
assignment_fsm: only include speech codec (choosen) on AoIP networks The field speech codec (choosen) in the ASSIGNMENT COMPLETE message has been specified with AoIP. Since sccp-lite has been specified before AoIP, we should not include the speech codec (choosen) field in the ASSIGNMENT COMPLETE, when osmo-bsc is connected to an sccp-lite based network. - make sure speech codec (choosen) is only included on AoIP based networks Change-Id: Ic7f28aacc953d96ebf3aa04f3e1810475458c1a8
2018-09-21 12:16:14 +00:00
/* Only AoIP networks include a speech codec (choosen) in the
* assignment complete message. */
if (gscon_is_aoip(conn)) {
/* Extrapolate speech codec from speech mode */
gsm0808_speech_codec_from_chan_type(&sc, perm_spch);
lchan_fsm: introduce lchan.activate.ch_mode_rate to allow tweaking lchan->activate.info.ch_mode_rate should remain unchanged, it is the immutable request data. However, for VAMOS, we will want to automatically see that the chan_mode is chosen correctly. As a first step, place a mutable ch_mode_rate copy at lchan->activate.ch_mode_rate, i.e. not in .activate.info, but in .activate. Use that everywhere. This is mostly a non-functional change, preparing for a subsequent patch that adds handling of VAMOS shadow lchans. As side effect of adding lchan_activate_set_ch_mode_rate_and_mr_config() after MODE_MODIF_ACK (enabling the voice stream after a channel mode modify), fix AMR config: the call to lchan_mr_config() was missing. Change-Id: Icc9cc7481b3984fdca34eef49ea91ad3388c06fe
2021-05-28 17:03:08 +00:00
sc.cfg = conn->lchan->activate.ch_mode_rate.s15_s0;
assignment_fsm: only include speech codec (choosen) on AoIP networks The field speech codec (choosen) in the ASSIGNMENT COMPLETE message has been specified with AoIP. Since sccp-lite has been specified before AoIP, we should not include the speech codec (choosen) field in the ASSIGNMENT COMPLETE, when osmo-bsc is connected to an sccp-lite based network. - make sure speech codec (choosen) is only included on AoIP based networks Change-Id: Ic7f28aacc953d96ebf3aa04f3e1810475458c1a8
2018-09-21 12:16:14 +00:00
sc_ptr = &sc;
}
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
}
LCLS: use libosmocore function to add status * use gsm0808_create_ass_compl2() to add BSS Status IE to Assignment Complete message * drop local helpers Depends-on: (libosmocore) I547c6b8707123aa8c1ef636db88908df112d90a4 Change-Id: I6916928391667cd9c345becf00e7c8561846c295 Related: OS#2487
2019-01-08 14:12:58 +00:00
resp = gsm0808_create_ass_compl2(lchan->abis_ip.ass_compl.rr_cause,
chosen_channel,
struct gsm_encr: store alg_id in human-readable format I find it weird that we store the A5 algorithm ID in a format that is used on the wire: N + 1 (valid for both A-bis and A interfaces). What confused me even more is that in some functions we print it as if it was in a normal, human-readable format. And this is also why one can see weird constructions like: if (lchan->encr.alg_id > ALG_A5_NR_TO_RSL(0)) { ... } Let's ensure that our internal structures use the A5/N format: alg_id=0: A5/0 (0x01 on the A-bis/A interface) alg_id=1: A5/1 (0x02 on the A-bis/A interface) alg_id=2: A5/2 (0x03 on the A-bis/A interface) ... alg_id=7: A5/7 (0x08 on the A-bis/A interface) so that we can print and compare the value of alg_id without using additional arithmetics. Let's also rename 'alg_id' to 'alg_a5_n' as it most clearly indicates which representation it is storing. This is how the above code snippet would look like: if (lchan->encr.alg_a5_n > 0) { ... } Change-Id: Ieb50c9a352cfa5481aebac2379e0a461663543ec
2022-03-13 22:01:53 +00:00
ALG_A5_NR_TO_BSSAP(lchan->encr.alg_a5_n), perm_spch,
LCLS: use libosmocore function to add status * use gsm0808_create_ass_compl2() to add BSS Status IE to Assignment Complete message * drop local helpers Depends-on: (libosmocore) I547c6b8707123aa8c1ef636db88908df112d90a4 Change-Id: I6916928391667cd9c345becf00e7c8561846c295 Related: OS#2487
2019-01-08 14:12:58 +00:00
addr_local_p, sc_ptr, NULL, lcls_get_status(conn));
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
if (!resp) {
assignment_fail(GSM0808_CAUSE_EQUIPMENT_FAILURE,
"Unable to compose Assignment Complete message");
return;
}
bssap: Feed Assign Complete with Osmux CID retrieved from MGW The CID is sent back to the MSC so it can configure its MGW to send Osmux frames to the BSC MGW using this CID. Change-Id: I77dfdd965ae828c39a9818669177aefd22bc02f2
2019-04-16 18:33:25 +00:00
if (gscon_is_aoip(conn) && conn->assignment.requires_voice_stream &&
conn->assignment.req.use_osmux)
_gsm0808_ass_compl_extend_osmux(resp, osmux_cid);
Use new stat item/ctr getter APIs Generated with following and similar spatch snippets: """ @@ expression E1, E2; @@ - &E2->ctr[E1] + rate_ctr_group_get_ctr(E2, E1) """ Change-Id: I0b43f922a595d694ac0aeda80107ef9bf4e755e7
2021-06-04 15:30:21 +00:00
rate_ctr_inc(rate_ctr_group_get_ctr(conn->sccp.msc->msc_ctrs, MSC_CTR_BSSMAP_TX_DT1_ASSIGMENT_COMPLETE));
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
rc = gscon_sigtran_send(conn, resp);
if (rc) {
assignment_fail(GSM0808_CAUSE_EQUIPMENT_FAILURE,
"Unable send Assignment Complete message: rc=%d %s",
rc, strerror(-rc));
return;
}
}
static void assignment_success(struct gsm_subscriber_connection *conn)
{
BS-11: Send proprietary MRPCI message after assignment + HO complete This is a non-standard RSL message that must be sent after each traffic channel has been established. Without it, any voice call will be disconnected within seconds. This is a hack; we need to store the subscribers classmark2 value and use it here. Change-Id: I6cb6d25e405aa888c4df4022d897330a6af9e946 Related: OS#2389
2020-08-01 15:03:31 +00:00
struct gsm_bts *bts;
lchan and assignment FSMs: make Channel Mode Modify more sane The Channel Mode Modify procedure is currently implemented for changing a TCH lchan from signalling to voice mode. For that, however, it is re-using (abusing) the channel activation structs and state transitions, and thus always implies activating a voice stream when the mode modification is done. I will add a Channel Mode Modify to enable VAMOS mode soon, so I require separate structs and state transitions which also work on an lchan that already has a voice stream established: a struct lchan_modify_info and LCHAN_EV_REQUEST_MODE_MODIFY, and dedicated assignment FSM state ASSIGNMENT_ST_WAIT_LCHAN_MODIFIED. For the part where a Channel Mode Modify enables a voice stream after switching from signalling to speech mode, still use the channel activation code path, but only once the mode modification is done. General improvements: - To ask for a mode modification, emit an FSM event that ensures a mode modify only happens when the lchan state allows it. - The new lchan_modify_info struct reflects only those parts that have an effect during a mode modification (before the lchan_activate_info was fully populated, many values not having an effect). - More accurate logging, indicating "Mode Modify" instead of "Channel Activation" A TTCN3 test for the Channel Mode Modify procedure is added in Idf4efaed986de0bbd2b663313e837352cc139f0f, and the test passes both before and after this patch is applied. Related: SYS#4895 Change-Id: I4986844f839b1c9672c61d916eb3d33d0042d747
2021-04-14 17:39:01 +00:00
bool lchan_changed = (conn->assignment.new_lchan != NULL);
BS-11: Send proprietary MRPCI message after assignment + HO complete This is a non-standard RSL message that must be sent after each traffic channel has been established. Without it, any voice call will be disconnected within seconds. This is a hack; we need to store the subscribers classmark2 value and use it here. Change-Id: I6cb6d25e405aa888c4df4022d897330a6af9e946 Related: OS#2389
2020-08-01 15:03:31 +00:00
lchan and assignment FSMs: make Channel Mode Modify more sane The Channel Mode Modify procedure is currently implemented for changing a TCH lchan from signalling to voice mode. For that, however, it is re-using (abusing) the channel activation structs and state transitions, and thus always implies activating a voice stream when the mode modification is done. I will add a Channel Mode Modify to enable VAMOS mode soon, so I require separate structs and state transitions which also work on an lchan that already has a voice stream established: a struct lchan_modify_info and LCHAN_EV_REQUEST_MODE_MODIFY, and dedicated assignment FSM state ASSIGNMENT_ST_WAIT_LCHAN_MODIFIED. For the part where a Channel Mode Modify enables a voice stream after switching from signalling to speech mode, still use the channel activation code path, but only once the mode modification is done. General improvements: - To ask for a mode modification, emit an FSM event that ensures a mode modify only happens when the lchan state allows it. - The new lchan_modify_info struct reflects only those parts that have an effect during a mode modification (before the lchan_activate_info was fully populated, many values not having an effect). - More accurate logging, indicating "Mode Modify" instead of "Channel Activation" A TTCN3 test for the Channel Mode Modify procedure is added in Idf4efaed986de0bbd2b663313e837352cc139f0f, and the test passes both before and after this patch is applied. Related: SYS#4895 Change-Id: I4986844f839b1c9672c61d916eb3d33d0042d747
2021-04-14 17:39:01 +00:00
/* Take on the new lchan. If there only was a Channel Mode Modify, then there is no new lchan to take on. */
if (lchan_changed) {
gscon_change_primary_lchan(conn, conn->assignment.new_lchan);
fix LCLS during Assignment: actually use new TCH lchan In assignment_success(), don't trigger lcls_apply_config() before the conn->lchan points at the actual new voice lchan. The entire LCLS code relies on conn->lchan. The assignment FSM wants to point conn->lchan to the new lchan only after it is sure that sending the Assignment Complete to the MSC was successful. However, a failure is a) very unlikely and b) if sending to the MSC fails we might as well tear the whole conn down anyway. Now, if sending to the MSC fails, release the lchan and the conn should clean itself up. Related: OS#1602 Change-Id: I8dd561d744d8081b5ac5ffa7635f17ac19bcda45
2018-09-24 15:38:04 +00:00
lchan and assignment FSMs: make Channel Mode Modify more sane The Channel Mode Modify procedure is currently implemented for changing a TCH lchan from signalling to voice mode. For that, however, it is re-using (abusing) the channel activation structs and state transitions, and thus always implies activating a voice stream when the mode modification is done. I will add a Channel Mode Modify to enable VAMOS mode soon, so I require separate structs and state transitions which also work on an lchan that already has a voice stream established: a struct lchan_modify_info and LCHAN_EV_REQUEST_MODE_MODIFY, and dedicated assignment FSM state ASSIGNMENT_ST_WAIT_LCHAN_MODIFIED. For the part where a Channel Mode Modify enables a voice stream after switching from signalling to speech mode, still use the channel activation code path, but only once the mode modification is done. General improvements: - To ask for a mode modification, emit an FSM event that ensures a mode modify only happens when the lchan state allows it. - The new lchan_modify_info struct reflects only those parts that have an effect during a mode modification (before the lchan_activate_info was fully populated, many values not having an effect). - More accurate logging, indicating "Mode Modify" instead of "Channel Activation" A TTCN3 test for the Channel Mode Modify procedure is added in Idf4efaed986de0bbd2b663313e837352cc139f0f, and the test passes both before and after this patch is applied. Related: SYS#4895 Change-Id: I4986844f839b1c9672c61d916eb3d33d0042d747
2021-04-14 17:39:01 +00:00
OSMO_ASSERT((bts = conn_get_bts(conn)) != NULL);
if (is_siemens_bts(bts) && ts_is_tch(conn->lchan->ts)) {
/* HACK: store the actual Classmark 2 LV from the subscriber and use it here! */
uint8_t cm2_lv[] = { 0x02, 0x00, 0x00 };
send_siemens_mrpci(conn->lchan, cm2_lv);
}
BS-11: Send proprietary MRPCI message after assignment + HO complete This is a non-standard RSL message that must be sent after each traffic channel has been established. Without it, any voice call will be disconnected within seconds. This is a hack; we need to store the subscribers classmark2 value and use it here. Change-Id: I6cb6d25e405aa888c4df4022d897330a6af9e946 Related: OS#2389
2020-08-01 15:03:31 +00:00
lchan and assignment FSMs: make Channel Mode Modify more sane The Channel Mode Modify procedure is currently implemented for changing a TCH lchan from signalling to voice mode. For that, however, it is re-using (abusing) the channel activation structs and state transitions, and thus always implies activating a voice stream when the mode modification is done. I will add a Channel Mode Modify to enable VAMOS mode soon, so I require separate structs and state transitions which also work on an lchan that already has a voice stream established: a struct lchan_modify_info and LCHAN_EV_REQUEST_MODE_MODIFY, and dedicated assignment FSM state ASSIGNMENT_ST_WAIT_LCHAN_MODIFIED. For the part where a Channel Mode Modify enables a voice stream after switching from signalling to speech mode, still use the channel activation code path, but only once the mode modification is done. General improvements: - To ask for a mode modification, emit an FSM event that ensures a mode modify only happens when the lchan state allows it. - The new lchan_modify_info struct reflects only those parts that have an effect during a mode modification (before the lchan_activate_info was fully populated, many values not having an effect). - More accurate logging, indicating "Mode Modify" instead of "Channel Activation" A TTCN3 test for the Channel Mode Modify procedure is added in Idf4efaed986de0bbd2b663313e837352cc139f0f, and the test passes both before and after this patch is applied. Related: SYS#4895 Change-Id: I4986844f839b1c9672c61d916eb3d33d0042d747
2021-04-14 17:39:01 +00:00
/* apply LCLS configuration (if any) */
lcls_apply_config(conn);
}
conn->assignment.new_lchan = NULL;
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
assignment_fsm: send BSSMAP response only after Assignment Request So far, only the MSC asked for Assignment via Assignment Request, which we answer with a BSSMAP Assignment Complete or Assignment Failure when done. When Assignment is triggered for any other reason (congestion resolution, VAMOS, VTY), we will not send any such messages to the MSC. Additional enum values will be added in subsequent commits: Id56a890106b93fcee67ac9401b890e7b63bba421 ASSIGN_FOR_CONGESTION_RESOLUTION If006f5caaf83b07675f57e5665cfa79328da55e6 ASSIGN_FOR_VTY Related: SYS#5315 OS#4940 Change-Id: Ie0cddbdb00abcec78e153f4ae6d04ce75080a111
2021-05-01 02:52:05 +00:00
if (conn->assignment.req.assign_for == ASSIGN_FOR_BSSMAP_REQ) {
send_assignment_complete(conn);
/* If something went wrong during send_assignment_complete(), the fi will be gone from
* error handling in there. Almost a success, but then again the whole thing failed. */
if (!conn->assignment.fi) {
/* The lchan was ready, and we failed to tell the MSC about it. By releasing this lchan,
* the conn will notice that its primary lchan is gone and should clean itself up. */
lchan_release(conn->lchan, true, true, RSL_ERR_EQUIPMENT_FAIL,
gscon_last_eutran_plmn(conn));
return;
}
fix LCLS during Assignment: actually use new TCH lchan In assignment_success(), don't trigger lcls_apply_config() before the conn->lchan points at the actual new voice lchan. The entire LCLS code relies on conn->lchan. The assignment FSM wants to point conn->lchan to the new lchan only after it is sure that sending the Assignment Complete to the MSC was successful. However, a failure is a) very unlikely and b) if sending to the MSC fails we might as well tear the whole conn down anyway. Now, if sending to the MSC fails, release the lchan and the conn should clean itself up. Related: OS#1602 Change-Id: I8dd561d744d8081b5ac5ffa7635f17ac19bcda45
2018-09-24 15:38:04 +00:00
}
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
assignment_fsm: always mark MGCP ci as completed When the assignment fails, we roll back the MSC side RTP endpoint at the MGW that may have been created before the failure occured. On success, we clear the mgcp_ci pointer so it is not rolled back. Always clear this, not only when lchan_changed == true. That is because a channel Mode Modify may also have added a voice stream that should retain the newly created RTP endpoint at the MGW. If no voice stream is involved, the pointer should already be NULL. There is no reason to have a condition for clearing this pointer. Just always clear it. This fixes all voice calls that modify a TCH lchan from signalling-only to voice mode. Before this patch, their MSC side RTP endpoint was DLCX'd right on assignment success. In particular, this fixes Emergency Calls (which usually get a TCH lchan assigned right from the start, and hence do a Mode Modify to add voice). Related: SYS#5916 Change-Id: I5ab10ee7fd9c5d7608e8a06893881d990943feed
2022-04-06 21:50:59 +00:00
/* Rembered this only for error handling: should assignment fail, assignment_reset() will release
* the MGW endpoint right away. If successful, the conn continues to use the endpoint. */
conn->assignment.created_ci_for_msc = NULL;
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
assignment_fsm: always update RTP info When the assignment succeeds, the assignment_request has taken effect and whatever it requested should now be in effect. Hence copy the new RTP information from the assignment_request to the conn's storage indicating what is currently used, always, not only when lchan_changed == true. The RTP information may have changed also from a Mode Modify where the lchan stays the same but changes its mode of use, e.g. from signalling to voice. When there is no RTP involved, this data is zero or empty, so there is no harm in copying it, always. Related: SYS#5916 Change-Id: I0788d1f013b8f820f559b6ed58a5f9bb8a02e0b4
2022-04-06 21:50:59 +00:00
/* New RTP information is now accepted. If there is no RTP stream, this information is zero / empty. Either way
* store the result of this assignment. */
conn->user_plane.msc_assigned_cic = conn->assignment.req.msc_assigned_cic;
osmo_strlcpy(conn->user_plane.msc_assigned_rtp_addr, conn->assignment.req.msc_rtp_addr,
sizeof(conn->user_plane.msc_assigned_rtp_addr));
conn->user_plane.msc_assigned_rtp_port = conn->assignment.req.msc_rtp_port;
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
fix assignment success counters: count *before* cleanup of fsm state Counting the Assignment success after osmo_fsm_inst_term() meant that we were counting a cleared out channel mode, which always yielded signalling mode. Count the Assignment success *before* terminating, so that we correctly count the successful assignment as speech mode. Related: SYS#4878 Related: Icb1386ec2ccd70eb3c026301b9b08ad7177278f7 (osmo-ttcn3-hacks) Change-Id: Ie9fcd1e86f27ecb2f11e2e8813faac365cb470b8
2021-11-29 15:25:34 +00:00
assignment_count_result(CTR_ASSIGNMENT_COMPLETED);
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
LOG_ASSIGNMENT(conn, LOGL_DEBUG, "Assignment successful\n");
osmo_fsm_inst_term(conn->assignment.fi, OSMO_FSM_TERM_REGULAR, 0);
}
static void assignment_fsm_update_id(struct gsm_subscriber_connection *conn)
{
lchan and assignment FSMs: make Channel Mode Modify more sane The Channel Mode Modify procedure is currently implemented for changing a TCH lchan from signalling to voice mode. For that, however, it is re-using (abusing) the channel activation structs and state transitions, and thus always implies activating a voice stream when the mode modification is done. I will add a Channel Mode Modify to enable VAMOS mode soon, so I require separate structs and state transitions which also work on an lchan that already has a voice stream established: a struct lchan_modify_info and LCHAN_EV_REQUEST_MODE_MODIFY, and dedicated assignment FSM state ASSIGNMENT_ST_WAIT_LCHAN_MODIFIED. For the part where a Channel Mode Modify enables a voice stream after switching from signalling to speech mode, still use the channel activation code path, but only once the mode modification is done. General improvements: - To ask for a mode modification, emit an FSM event that ensures a mode modify only happens when the lchan state allows it. - The new lchan_modify_info struct reflects only those parts that have an effect during a mode modification (before the lchan_activate_info was fully populated, many values not having an effect). - More accurate logging, indicating "Mode Modify" instead of "Channel Activation" A TTCN3 test for the Channel Mode Modify procedure is added in Idf4efaed986de0bbd2b663313e837352cc139f0f, and the test passes both before and after this patch is applied. Related: SYS#4895 Change-Id: I4986844f839b1c9672c61d916eb3d33d0042d747
2021-04-14 17:39:01 +00:00
/* Assignment can do a new channel activation, in which case new_lchan points at the new lchan.
* Or assignment can Channel Mode Modify the already used lchan, in which case new_lchan == NULL. */
struct gsm_lchan *new_lchan = (conn->assignment.new_lchan ? : conn->lchan);
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
if (!new_lchan) {
osmo_fsm_inst_update_id(conn->assignment.fi, conn->fi->id);
return;
}
update the lchan name to always reflect VAMOS shadowness Change gsm_lchan_name_compute() to a function that in-place updates the lchan->name. That allows calling it numerous times with the talloc handled internally. Rename it to lchan_update_name(). Add 'shadow' to lchan_update_name() and lchan_fsm_update_id() for VAMOS shadow lchans, and also print the lchan index that it is a shadow for, instead of the index in the lchan array. When set_pchan_is() updates the VAMOSness of the lchans, call lchan_fsm_update_id(). From lchan_fsm_update_id() also call lchan_update_name(). This is a bit convoluted for legacy reasons. There are utility programs and C tests using bts_trx.c but not lchan_fsm.c. lchan_update_name() lives in gsm_data.c for that reason. This patch calls lchan_update_name() from lchan_fsm_update_id() and not vice versa to avoid having to add stubbed lchan_fsm_update_id() functions to all utility programs and C tests. We can't easily unify the lchan->name and lchan->fi->id without lots of refactoring rippling through all those little utility programs and C tests. Change-Id: I7c2bae3b895a91f1b99b4147ecc0e3009cb7439a
2021-05-29 23:08:22 +00:00
osmo_fsm_inst_update_id_f(conn->assignment.fi, "%s_%u-%u-%u-%s%s%s-%s%u",
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
conn->fi->id,
new_lchan->ts->trx->bts->nr, new_lchan->ts->trx->nr, new_lchan->ts->nr,
gsm_pchan_id(new_lchan->ts->pchan_on_init),
(new_lchan->ts->pchan_on_init == new_lchan->ts->pchan_is)? "" : "as",
(new_lchan->ts->pchan_on_init == new_lchan->ts->pchan_is)? ""
: gsm_pchan_id(new_lchan->ts->pchan_is),
update the lchan name to always reflect VAMOS shadowness Change gsm_lchan_name_compute() to a function that in-place updates the lchan->name. That allows calling it numerous times with the talloc handled internally. Rename it to lchan_update_name(). Add 'shadow' to lchan_update_name() and lchan_fsm_update_id() for VAMOS shadow lchans, and also print the lchan index that it is a shadow for, instead of the index in the lchan array. When set_pchan_is() updates the VAMOSness of the lchans, call lchan_fsm_update_id(). From lchan_fsm_update_id() also call lchan_update_name(). This is a bit convoluted for legacy reasons. There are utility programs and C tests using bts_trx.c but not lchan_fsm.c. lchan_update_name() lives in gsm_data.c for that reason. This patch calls lchan_update_name() from lchan_fsm_update_id() and not vice versa to avoid having to add stubbed lchan_fsm_update_id() functions to all utility programs and C tests. We can't easily unify the lchan->name and lchan->fi->id without lots of refactoring rippling through all those little utility programs and C tests. Change-Id: I7c2bae3b895a91f1b99b4147ecc0e3009cb7439a
2021-05-29 23:08:22 +00:00
new_lchan->vamos.is_secondary ? "shadow" : "",
new_lchan->nr - (new_lchan->vamos.is_secondary ? new_lchan->ts->max_primary_lchans : 0));
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
}
assignment_fsm: fix channel allocator preferences When the MSC allocates a channel through the ASSIGNMENT REQUEST, it may ask for a TCH/H and a TCH/F at the same time and tell which of the two types it prefers. The process of channel allocation currently selects, based on the BTS, MSC and MS capabilites exactly one apropriate codec/rate (e.g. TCH/H) and then tries to allocate it. If that allocation fails, there is no way to try the second choice and the assignment fails. For example: The MSC asks for TCH/F and TCH/H, prefering TCH/F, then the channel allocator will try TCH/F and if it fails (all TCH/F are currently in use), then TCH/H is never tried. Since the BSC currently only trys the first best codec/rate that is supported it also ignores the preference. Lets fix those problems by including the preference information and both possible codec/rate settings into the channel allocation decision. Change-Id: I5239e05c1cfbcb8af28f43373a58fa6c2d216c51 Related: OS#3503
2019-01-22 10:29:06 +00:00
static bool lchan_type_compat_with_mode(enum gsm_chan_t type, const struct channel_mode_and_rate *ch_mode_rate)
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
{
assignment_fsm: fix channel allocator preferences When the MSC allocates a channel through the ASSIGNMENT REQUEST, it may ask for a TCH/H and a TCH/F at the same time and tell which of the two types it prefers. The process of channel allocation currently selects, based on the BTS, MSC and MS capabilites exactly one apropriate codec/rate (e.g. TCH/H) and then tries to allocate it. If that allocation fails, there is no way to try the second choice and the assignment fails. For example: The MSC asks for TCH/F and TCH/H, prefering TCH/F, then the channel allocator will try TCH/F and if it fails (all TCH/F are currently in use), then TCH/H is never tried. Since the BSC currently only trys the first best codec/rate that is supported it also ignores the preference. Lets fix those problems by including the preference information and both possible codec/rate settings into the channel allocation decision. Change-Id: I5239e05c1cfbcb8af28f43373a58fa6c2d216c51 Related: OS#3503
2019-01-22 10:29:06 +00:00
enum gsm48_chan_mode chan_mode = ch_mode_rate->chan_mode;
assignment_fsm: Properly support assigning signalling mode TCH/x To support the 3 possible preferences, the changes needed were: - Replace 'full_rate' bool with a 3 option enum to represent the channels types for signalling - Switch from _pref/_alt to using an array sorted in preference order Originally merged as Change-Id I4c7499c8c866ea3ff7b1327edb3615d003d927d3, reverted because the change broke voice calls. Re-submitting with the fix: don't forget to set conn->assignment.requires_voice_stream. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I7513d2cbe8b695ba6f031ad11560c63a6535cf2d
2019-03-07 15:32:02 +00:00
enum channel_rate chan_rate = ch_mode_rate->chan_rate;
assignment_fsm: fix channel allocator preferences When the MSC allocates a channel through the ASSIGNMENT REQUEST, it may ask for a TCH/H and a TCH/F at the same time and tell which of the two types it prefers. The process of channel allocation currently selects, based on the BTS, MSC and MS capabilites exactly one apropriate codec/rate (e.g. TCH/H) and then tries to allocate it. If that allocation fails, there is no way to try the second choice and the assignment fails. For example: The MSC asks for TCH/F and TCH/H, prefering TCH/F, then the channel allocator will try TCH/F and if it fails (all TCH/F are currently in use), then TCH/H is never tried. Since the BSC currently only trys the first best codec/rate that is supported it also ignores the preference. Lets fix those problems by including the preference information and both possible codec/rate settings into the channel allocation decision. Change-Id: I5239e05c1cfbcb8af28f43373a58fa6c2d216c51 Related: OS#3503
2019-01-22 10:29:06 +00:00
ensure chan_mode comparisons in non-VAMOS mode Both VAMOS- and non-VAMOS speech modes should result in indentical voice handling. So make sure that all chan_modes are converted to non-vamos before comparing / evaluating in switch statements. Change-Id: I791e7966b1f8eaa3299a8a46abeb313cf5136e0b
2021-05-28 17:18:11 +00:00
switch (gsm48_chan_mode_to_non_vamos(chan_mode)) {
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
case GSM48_CMODE_SIGN:
switch (type) {
assignment_fsm: Properly support assigning signalling mode TCH/x To support the 3 possible preferences, the changes needed were: - Replace 'full_rate' bool with a 3 option enum to represent the channels types for signalling - Switch from _pref/_alt to using an array sorted in preference order Originally merged as Change-Id I4c7499c8c866ea3ff7b1327edb3615d003d927d3, reverted because the change broke voice calls. Re-submitting with the fix: don't forget to set conn->assignment.requires_voice_stream. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I7513d2cbe8b695ba6f031ad11560c63a6535cf2d
2019-03-07 15:32:02 +00:00
case GSM_LCHAN_TCH_F: return chan_rate == CH_RATE_FULL;
case GSM_LCHAN_TCH_H: return chan_rate == CH_RATE_HALF;
case GSM_LCHAN_SDCCH: return chan_rate == CH_RATE_SDCCH;
default: return false;
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
}
case GSM48_CMODE_SPEECH_V1:
case GSM48_CMODE_SPEECH_AMR:
case GSM48_CMODE_DATA_3k6:
case GSM48_CMODE_DATA_6k0:
/* these services can all run on TCH/H, but we may have
assignment_fsm: Properly support assigning signalling mode TCH/x To support the 3 possible preferences, the changes needed were: - Replace 'full_rate' bool with a 3 option enum to represent the channels types for signalling - Switch from _pref/_alt to using an array sorted in preference order Originally merged as Change-Id I4c7499c8c866ea3ff7b1327edb3615d003d927d3, reverted because the change broke voice calls. Re-submitting with the fix: don't forget to set conn->assignment.requires_voice_stream. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I7513d2cbe8b695ba6f031ad11560c63a6535cf2d
2019-03-07 15:32:02 +00:00
* an explicit override by the 'chan_rate' argument */
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
switch (type) {
case GSM_LCHAN_TCH_F:
assignment_fsm: Properly support assigning signalling mode TCH/x To support the 3 possible preferences, the changes needed were: - Replace 'full_rate' bool with a 3 option enum to represent the channels types for signalling - Switch from _pref/_alt to using an array sorted in preference order Originally merged as Change-Id I4c7499c8c866ea3ff7b1327edb3615d003d927d3, reverted because the change broke voice calls. Re-submitting with the fix: don't forget to set conn->assignment.requires_voice_stream. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I7513d2cbe8b695ba6f031ad11560c63a6535cf2d
2019-03-07 15:32:02 +00:00
return chan_rate == CH_RATE_FULL;
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
case GSM_LCHAN_TCH_H:
assignment_fsm: Properly support assigning signalling mode TCH/x To support the 3 possible preferences, the changes needed were: - Replace 'full_rate' bool with a 3 option enum to represent the channels types for signalling - Switch from _pref/_alt to using an array sorted in preference order Originally merged as Change-Id I4c7499c8c866ea3ff7b1327edb3615d003d927d3, reverted because the change broke voice calls. Re-submitting with the fix: don't forget to set conn->assignment.requires_voice_stream. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I7513d2cbe8b695ba6f031ad11560c63a6535cf2d
2019-03-07 15:32:02 +00:00
return chan_rate == CH_RATE_HALF;
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
default:
return false;
}
case GSM48_CMODE_DATA_12k0:
case GSM48_CMODE_DATA_14k5:
case GSM48_CMODE_SPEECH_EFR:
/* these services all explicitly require a TCH/F */
return type == GSM_LCHAN_TCH_F;
default:
return false;
}
}
void assignment_fsm_init()
{
OSMO_ASSERT(osmo_fsm_register(&assignment_fsm) == 0);
}
assignment_fsm: fix channel allocator preferences When the MSC allocates a channel through the ASSIGNMENT REQUEST, it may ask for a TCH/H and a TCH/F at the same time and tell which of the two types it prefers. The process of channel allocation currently selects, based on the BTS, MSC and MS capabilites exactly one apropriate codec/rate (e.g. TCH/H) and then tries to allocate it. If that allocation fails, there is no way to try the second choice and the assignment fails. For example: The MSC asks for TCH/F and TCH/H, prefering TCH/F, then the channel allocator will try TCH/F and if it fails (all TCH/F are currently in use), then TCH/H is never tried. Since the BSC currently only trys the first best codec/rate that is supported it also ignores the preference. Lets fix those problems by including the preference information and both possible codec/rate settings into the channel allocation decision. Change-Id: I5239e05c1cfbcb8af28f43373a58fa6c2d216c51 Related: OS#3503
2019-01-22 10:29:06 +00:00
static int check_requires_voice(bool *requires_voice, enum gsm48_chan_mode chan_mode)
{
*requires_voice = false;
ensure chan_mode comparisons in non-VAMOS mode Both VAMOS- and non-VAMOS speech modes should result in indentical voice handling. So make sure that all chan_modes are converted to non-vamos before comparing / evaluating in switch statements. Change-Id: I791e7966b1f8eaa3299a8a46abeb313cf5136e0b
2021-05-28 17:18:11 +00:00
switch (gsm48_chan_mode_to_non_vamos(chan_mode)) {
assignment_fsm: fix channel allocator preferences When the MSC allocates a channel through the ASSIGNMENT REQUEST, it may ask for a TCH/H and a TCH/F at the same time and tell which of the two types it prefers. The process of channel allocation currently selects, based on the BTS, MSC and MS capabilites exactly one apropriate codec/rate (e.g. TCH/H) and then tries to allocate it. If that allocation fails, there is no way to try the second choice and the assignment fails. For example: The MSC asks for TCH/F and TCH/H, prefering TCH/F, then the channel allocator will try TCH/F and if it fails (all TCH/F are currently in use), then TCH/H is never tried. Since the BSC currently only trys the first best codec/rate that is supported it also ignores the preference. Lets fix those problems by including the preference information and both possible codec/rate settings into the channel allocation decision. Change-Id: I5239e05c1cfbcb8af28f43373a58fa6c2d216c51 Related: OS#3503
2019-01-22 10:29:06 +00:00
case GSM48_CMODE_SPEECH_V1:
case GSM48_CMODE_SPEECH_EFR:
case GSM48_CMODE_SPEECH_AMR:
*requires_voice = true;
break;
case GSM48_CMODE_SIGN:
*requires_voice = false;
break;
default:
return -EINVAL;
}
return 0;
}
/* Check if the incoming assignment requests requires a voice stream or not,
* we will look at the preferred and the alternate channel mode and also make
* sure that both are consistent. */
static int check_requires_voice_stream(struct gsm_subscriber_connection *conn)
{
assignment_fsm: Properly support assigning signalling mode TCH/x To support the 3 possible preferences, the changes needed were: - Replace 'full_rate' bool with a 3 option enum to represent the channels types for signalling - Switch from _pref/_alt to using an array sorted in preference order Originally merged as Change-Id I4c7499c8c866ea3ff7b1327edb3615d003d927d3, reverted because the change broke voice calls. Re-submitting with the fix: don't forget to set conn->assignment.requires_voice_stream. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I7513d2cbe8b695ba6f031ad11560c63a6535cf2d
2019-03-07 15:32:02 +00:00
bool requires_voice_pref = false, requires_voice_alt;
assignment_fsm: fix channel allocator preferences When the MSC allocates a channel through the ASSIGNMENT REQUEST, it may ask for a TCH/H and a TCH/F at the same time and tell which of the two types it prefers. The process of channel allocation currently selects, based on the BTS, MSC and MS capabilites exactly one apropriate codec/rate (e.g. TCH/H) and then tries to allocate it. If that allocation fails, there is no way to try the second choice and the assignment fails. For example: The MSC asks for TCH/F and TCH/H, prefering TCH/F, then the channel allocator will try TCH/F and if it fails (all TCH/F are currently in use), then TCH/H is never tried. Since the BSC currently only trys the first best codec/rate that is supported it also ignores the preference. Lets fix those problems by including the preference information and both possible codec/rate settings into the channel allocation decision. Change-Id: I5239e05c1cfbcb8af28f43373a58fa6c2d216c51 Related: OS#3503
2019-01-22 10:29:06 +00:00
struct assignment_request *req = &conn->assignment.req;
struct osmo_fsm_inst *fi = conn->fi;
assignment_fsm: Properly support assigning signalling mode TCH/x To support the 3 possible preferences, the changes needed were: - Replace 'full_rate' bool with a 3 option enum to represent the channels types for signalling - Switch from _pref/_alt to using an array sorted in preference order Originally merged as Change-Id I4c7499c8c866ea3ff7b1327edb3615d003d927d3, reverted because the change broke voice calls. Re-submitting with the fix: don't forget to set conn->assignment.requires_voice_stream. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I7513d2cbe8b695ba6f031ad11560c63a6535cf2d
2019-03-07 15:32:02 +00:00
int i, rc;
assignment_fsm: fix channel allocator preferences When the MSC allocates a channel through the ASSIGNMENT REQUEST, it may ask for a TCH/H and a TCH/F at the same time and tell which of the two types it prefers. The process of channel allocation currently selects, based on the BTS, MSC and MS capabilites exactly one apropriate codec/rate (e.g. TCH/H) and then tries to allocate it. If that allocation fails, there is no way to try the second choice and the assignment fails. For example: The MSC asks for TCH/F and TCH/H, prefering TCH/F, then the channel allocator will try TCH/F and if it fails (all TCH/F are currently in use), then TCH/H is never tried. Since the BSC currently only trys the first best codec/rate that is supported it also ignores the preference. Lets fix those problems by including the preference information and both possible codec/rate settings into the channel allocation decision. Change-Id: I5239e05c1cfbcb8af28f43373a58fa6c2d216c51 Related: OS#3503
2019-01-22 10:29:06 +00:00
/* When the assignment request indicates that there is an alternate
* rate available (e.g. "Full or Half rate channel, Half rate
* preferred..."), then both must be either voice or either signalling,
* a mismatch is not permitted */
assignment_fsm: Properly support assigning signalling mode TCH/x To support the 3 possible preferences, the changes needed were: - Replace 'full_rate' bool with a 3 option enum to represent the channels types for signalling - Switch from _pref/_alt to using an array sorted in preference order Originally merged as Change-Id I4c7499c8c866ea3ff7b1327edb3615d003d927d3, reverted because the change broke voice calls. Re-submitting with the fix: don't forget to set conn->assignment.requires_voice_stream. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I7513d2cbe8b695ba6f031ad11560c63a6535cf2d
2019-03-07 15:32:02 +00:00
for (i = 0; i < req->n_ch_mode_rate; i++) {
make sure channel mode and s15_s0 are updated only after an ACK I noticed during testing that an lchan used as TCH/F in fact still had its channel mode set to Signalling -- because on Assignment, the Speech mode used to be placed in the *previous* lchan and the new lchan was never updated after the Activ ACK. This is unbearable confusion which I complained about numerous times, so far mostly for cosmetic reasons. But implementing re-assignment properly actually requires this to be cleaned up. Keep all volatile chan mode settings in lchan->activate.* or lchan->modify.*, and only update lchan->* members when an ACK has been received for those settings. So a failed request keeps a sane state. Make sure that those settings are in fact updated in the proper lchan, upon an ACK, so that subsequent re-assignment or mode-modify know the accurate lchan state. Related are upcoming patches that sort out the AMR multirate configuration in a similar fashion, see Iebac2dc26412d877e5364f90d6f2ed7a7952351e Ia7519d2fa9e7f0b61b222d27d077bde4660c40b9 Ie57f9d0e3912632903d9740291225bfd1634ed47. Related: SYS#5315 OS#4940 OS#3787 OS#3833 Change-Id: Ie0da36124d73efc28a8809b63d7c96e2167fc412
2021-04-27 23:17:14 +00:00
rc = check_requires_voice(&requires_voice_alt, req->ch_mode_rate_list[i].chan_mode);
assignment_fsm: Properly support assigning signalling mode TCH/x To support the 3 possible preferences, the changes needed were: - Replace 'full_rate' bool with a 3 option enum to represent the channels types for signalling - Switch from _pref/_alt to using an array sorted in preference order Originally merged as Change-Id I4c7499c8c866ea3ff7b1327edb3615d003d927d3, reverted because the change broke voice calls. Re-submitting with the fix: don't forget to set conn->assignment.requires_voice_stream. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I7513d2cbe8b695ba6f031ad11560c63a6535cf2d
2019-03-07 15:32:02 +00:00
if (rc < 0) {
assignment_fail(GSM0808_CAUSE_REQ_CODEC_TYPE_OR_CONFIG_NOT_SUPP,
"Channel mode not supported (prev level %d): %s", i,
make sure channel mode and s15_s0 are updated only after an ACK I noticed during testing that an lchan used as TCH/F in fact still had its channel mode set to Signalling -- because on Assignment, the Speech mode used to be placed in the *previous* lchan and the new lchan was never updated after the Activ ACK. This is unbearable confusion which I complained about numerous times, so far mostly for cosmetic reasons. But implementing re-assignment properly actually requires this to be cleaned up. Keep all volatile chan mode settings in lchan->activate.* or lchan->modify.*, and only update lchan->* members when an ACK has been received for those settings. So a failed request keeps a sane state. Make sure that those settings are in fact updated in the proper lchan, upon an ACK, so that subsequent re-assignment or mode-modify know the accurate lchan state. Related are upcoming patches that sort out the AMR multirate configuration in a similar fashion, see Iebac2dc26412d877e5364f90d6f2ed7a7952351e Ia7519d2fa9e7f0b61b222d27d077bde4660c40b9 Ie57f9d0e3912632903d9740291225bfd1634ed47. Related: SYS#5315 OS#4940 OS#3787 OS#3833 Change-Id: Ie0da36124d73efc28a8809b63d7c96e2167fc412
2021-04-27 23:17:14 +00:00
gsm48_chan_mode_name(req->ch_mode_rate_list[i].chan_mode));
assignment_fsm: Properly support assigning signalling mode TCH/x To support the 3 possible preferences, the changes needed were: - Replace 'full_rate' bool with a 3 option enum to represent the channels types for signalling - Switch from _pref/_alt to using an array sorted in preference order Originally merged as Change-Id I4c7499c8c866ea3ff7b1327edb3615d003d927d3, reverted because the change broke voice calls. Re-submitting with the fix: don't forget to set conn->assignment.requires_voice_stream. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I7513d2cbe8b695ba6f031ad11560c63a6535cf2d
2019-03-07 15:32:02 +00:00
return -EINVAL;
}
assignment_fsm: fix channel allocator preferences When the MSC allocates a channel through the ASSIGNMENT REQUEST, it may ask for a TCH/H and a TCH/F at the same time and tell which of the two types it prefers. The process of channel allocation currently selects, based on the BTS, MSC and MS capabilites exactly one apropriate codec/rate (e.g. TCH/H) and then tries to allocate it. If that allocation fails, there is no way to try the second choice and the assignment fails. For example: The MSC asks for TCH/F and TCH/H, prefering TCH/F, then the channel allocator will try TCH/F and if it fails (all TCH/F are currently in use), then TCH/H is never tried. Since the BSC currently only trys the first best codec/rate that is supported it also ignores the preference. Lets fix those problems by including the preference information and both possible codec/rate settings into the channel allocation decision. Change-Id: I5239e05c1cfbcb8af28f43373a58fa6c2d216c51 Related: OS#3503
2019-01-22 10:29:06 +00:00
assignment_fsm: Properly support assigning signalling mode TCH/x To support the 3 possible preferences, the changes needed were: - Replace 'full_rate' bool with a 3 option enum to represent the channels types for signalling - Switch from _pref/_alt to using an array sorted in preference order Originally merged as Change-Id I4c7499c8c866ea3ff7b1327edb3615d003d927d3, reverted because the change broke voice calls. Re-submitting with the fix: don't forget to set conn->assignment.requires_voice_stream. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I7513d2cbe8b695ba6f031ad11560c63a6535cf2d
2019-03-07 15:32:02 +00:00
if (i==0)
requires_voice_pref = requires_voice_alt;
else if (requires_voice_alt != requires_voice_pref) {
assignment_fail(GSM0808_CAUSE_REQ_CODEC_TYPE_OR_CONFIG_NOT_SUPP,
assignment_fsm: tweak error log msg for mixed modes Change-Id: Ic42b141012215f428066d739a2099b217f4e5d41
2021-04-28 15:19:10 +00:00
"Requested a mix of Signalling and non-Signalling channel modes: %s != %s",
make sure channel mode and s15_s0 are updated only after an ACK I noticed during testing that an lchan used as TCH/F in fact still had its channel mode set to Signalling -- because on Assignment, the Speech mode used to be placed in the *previous* lchan and the new lchan was never updated after the Activ ACK. This is unbearable confusion which I complained about numerous times, so far mostly for cosmetic reasons. But implementing re-assignment properly actually requires this to be cleaned up. Keep all volatile chan mode settings in lchan->activate.* or lchan->modify.*, and only update lchan->* members when an ACK has been received for those settings. So a failed request keeps a sane state. Make sure that those settings are in fact updated in the proper lchan, upon an ACK, so that subsequent re-assignment or mode-modify know the accurate lchan state. Related are upcoming patches that sort out the AMR multirate configuration in a similar fashion, see Iebac2dc26412d877e5364f90d6f2ed7a7952351e Ia7519d2fa9e7f0b61b222d27d077bde4660c40b9 Ie57f9d0e3912632903d9740291225bfd1634ed47. Related: SYS#5315 OS#4940 OS#3787 OS#3833 Change-Id: Ie0da36124d73efc28a8809b63d7c96e2167fc412
2021-04-27 23:17:14 +00:00
gsm48_chan_mode_name(req->ch_mode_rate_list[0].chan_mode),
gsm48_chan_mode_name(req->ch_mode_rate_list[i].chan_mode));
assignment_fsm: Properly support assigning signalling mode TCH/x To support the 3 possible preferences, the changes needed were: - Replace 'full_rate' bool with a 3 option enum to represent the channels types for signalling - Switch from _pref/_alt to using an array sorted in preference order Originally merged as Change-Id I4c7499c8c866ea3ff7b1327edb3615d003d927d3, reverted because the change broke voice calls. Re-submitting with the fix: don't forget to set conn->assignment.requires_voice_stream. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I7513d2cbe8b695ba6f031ad11560c63a6535cf2d
2019-03-07 15:32:02 +00:00
return -EINVAL;
}
assignment_fsm: fix channel allocator preferences When the MSC allocates a channel through the ASSIGNMENT REQUEST, it may ask for a TCH/H and a TCH/F at the same time and tell which of the two types it prefers. The process of channel allocation currently selects, based on the BTS, MSC and MS capabilites exactly one apropriate codec/rate (e.g. TCH/H) and then tries to allocate it. If that allocation fails, there is no way to try the second choice and the assignment fails. For example: The MSC asks for TCH/F and TCH/H, prefering TCH/F, then the channel allocator will try TCH/F and if it fails (all TCH/F are currently in use), then TCH/H is never tried. Since the BSC currently only trys the first best codec/rate that is supported it also ignores the preference. Lets fix those problems by including the preference information and both possible codec/rate settings into the channel allocation decision. Change-Id: I5239e05c1cfbcb8af28f43373a58fa6c2d216c51 Related: OS#3503
2019-01-22 10:29:06 +00:00
}
assignment_fsm: Properly support assigning signalling mode TCH/x To support the 3 possible preferences, the changes needed were: - Replace 'full_rate' bool with a 3 option enum to represent the channels types for signalling - Switch from _pref/_alt to using an array sorted in preference order Originally merged as Change-Id I4c7499c8c866ea3ff7b1327edb3615d003d927d3, reverted because the change broke voice calls. Re-submitting with the fix: don't forget to set conn->assignment.requires_voice_stream. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I7513d2cbe8b695ba6f031ad11560c63a6535cf2d
2019-03-07 15:32:02 +00:00
conn->assignment.requires_voice_stream = requires_voice_pref;
assignment_fsm: fix channel allocator preferences When the MSC allocates a channel through the ASSIGNMENT REQUEST, it may ask for a TCH/H and a TCH/F at the same time and tell which of the two types it prefers. The process of channel allocation currently selects, based on the BTS, MSC and MS capabilites exactly one apropriate codec/rate (e.g. TCH/H) and then tries to allocate it. If that allocation fails, there is no way to try the second choice and the assignment fails. For example: The MSC asks for TCH/F and TCH/H, prefering TCH/F, then the channel allocator will try TCH/F and if it fails (all TCH/F are currently in use), then TCH/H is never tried. Since the BSC currently only trys the first best codec/rate that is supported it also ignores the preference. Lets fix those problems by including the preference information and both possible codec/rate settings into the channel allocation decision. Change-Id: I5239e05c1cfbcb8af28f43373a58fa6c2d216c51 Related: OS#3503
2019-01-22 10:29:06 +00:00
return 0;
}
lchan_fsm: make rsl mode-modify working again osmo-nitb supports the modification of an lchan if the lchan is compatible but in the wrong mode. This feature was dropped in the transition to AoIP/bsc-split. However, osmo-bsc still has code to generate and parse the messeages, but the FSMs do not support a mode modify yetm Lets add handling for mode-modify to the lchan_fsm and assignment_fsm in order to support mode modify again Change-Id: I2c5a283b1ee33745cc1fcfcc09a0f9382224e2eb Related: OS#4549
2020-08-21 18:47:49 +00:00
/* Decide if we should re-use an existing lchan. For this we check if the
* current lchan is compatible with one of the requested modes. */
assignment_fsm: fix channel allocator preferences When the MSC allocates a channel through the ASSIGNMENT REQUEST, it may ask for a TCH/H and a TCH/F at the same time and tell which of the two types it prefers. The process of channel allocation currently selects, based on the BTS, MSC and MS capabilites exactly one apropriate codec/rate (e.g. TCH/H) and then tries to allocate it. If that allocation fails, there is no way to try the second choice and the assignment fails. For example: The MSC asks for TCH/F and TCH/H, prefering TCH/F, then the channel allocator will try TCH/F and if it fails (all TCH/F are currently in use), then TCH/H is never tried. Since the BSC currently only trys the first best codec/rate that is supported it also ignores the preference. Lets fix those problems by including the preference information and both possible codec/rate settings into the channel allocation decision. Change-Id: I5239e05c1cfbcb8af28f43373a58fa6c2d216c51 Related: OS#3503
2019-01-22 10:29:06 +00:00
static bool reuse_existing_lchan(struct gsm_subscriber_connection *conn)
{
struct assignment_request *req = &conn->assignment.req;
assignment_fsm: Properly support assigning signalling mode TCH/x To support the 3 possible preferences, the changes needed were: - Replace 'full_rate' bool with a 3 option enum to represent the channels types for signalling - Switch from _pref/_alt to using an array sorted in preference order Originally merged as Change-Id I4c7499c8c866ea3ff7b1327edb3615d003d927d3, reverted because the change broke voice calls. Re-submitting with the fix: don't forget to set conn->assignment.requires_voice_stream. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I7513d2cbe8b695ba6f031ad11560c63a6535cf2d
2019-03-07 15:32:02 +00:00
int i;
assignment_fsm: fix channel allocator preferences When the MSC allocates a channel through the ASSIGNMENT REQUEST, it may ask for a TCH/H and a TCH/F at the same time and tell which of the two types it prefers. The process of channel allocation currently selects, based on the BTS, MSC and MS capabilites exactly one apropriate codec/rate (e.g. TCH/H) and then tries to allocate it. If that allocation fails, there is no way to try the second choice and the assignment fails. For example: The MSC asks for TCH/F and TCH/H, prefering TCH/F, then the channel allocator will try TCH/F and if it fails (all TCH/F are currently in use), then TCH/H is never tried. Since the BSC currently only trys the first best codec/rate that is supported it also ignores the preference. Lets fix those problems by including the preference information and both possible codec/rate settings into the channel allocation decision. Change-Id: I5239e05c1cfbcb8af28f43373a58fa6c2d216c51 Related: OS#3503
2019-01-22 10:29:06 +00:00
if (!conn->lchan)
return false;
/* Check if the currently existing lchan is compatible with the
* preferred rate/codec. */
cosmetic scoping in reuse_existing_lchan() cosmetic prep for Ie0da36124d73efc28a8809b63d7c96e2167fc412 Related: SYS#5315 OS#4940 Change-Id: I0d1000cf108b0dfdc690f7a59d9bdf40de9043d2
2021-04-28 14:55:27 +00:00
for (i = 0; i < req->n_ch_mode_rate; i++) {
make sure channel mode and s15_s0 are updated only after an ACK I noticed during testing that an lchan used as TCH/F in fact still had its channel mode set to Signalling -- because on Assignment, the Speech mode used to be placed in the *previous* lchan and the new lchan was never updated after the Activ ACK. This is unbearable confusion which I complained about numerous times, so far mostly for cosmetic reasons. But implementing re-assignment properly actually requires this to be cleaned up. Keep all volatile chan mode settings in lchan->activate.* or lchan->modify.*, and only update lchan->* members when an ACK has been received for those settings. So a failed request keeps a sane state. Make sure that those settings are in fact updated in the proper lchan, upon an ACK, so that subsequent re-assignment or mode-modify know the accurate lchan state. Related are upcoming patches that sort out the AMR multirate configuration in a similar fashion, see Iebac2dc26412d877e5364f90d6f2ed7a7952351e Ia7519d2fa9e7f0b61b222d27d077bde4660c40b9 Ie57f9d0e3912632903d9740291225bfd1634ed47. Related: SYS#5315 OS#4940 OS#3787 OS#3833 Change-Id: Ie0da36124d73efc28a8809b63d7c96e2167fc412
2021-04-27 23:17:14 +00:00
if (!lchan_type_compat_with_mode(conn->lchan->type, &req->ch_mode_rate_list[i]))
cosmetic scoping in reuse_existing_lchan() cosmetic prep for Ie0da36124d73efc28a8809b63d7c96e2167fc412 Related: SYS#5315 OS#4940 Change-Id: I0d1000cf108b0dfdc690f7a59d9bdf40de9043d2
2021-04-28 14:55:27 +00:00
continue;
make sure channel mode and s15_s0 are updated only after an ACK I noticed during testing that an lchan used as TCH/F in fact still had its channel mode set to Signalling -- because on Assignment, the Speech mode used to be placed in the *previous* lchan and the new lchan was never updated after the Activ ACK. This is unbearable confusion which I complained about numerous times, so far mostly for cosmetic reasons. But implementing re-assignment properly actually requires this to be cleaned up. Keep all volatile chan mode settings in lchan->activate.* or lchan->modify.*, and only update lchan->* members when an ACK has been received for those settings. So a failed request keeps a sane state. Make sure that those settings are in fact updated in the proper lchan, upon an ACK, so that subsequent re-assignment or mode-modify know the accurate lchan state. Related are upcoming patches that sort out the AMR multirate configuration in a similar fashion, see Iebac2dc26412d877e5364f90d6f2ed7a7952351e Ia7519d2fa9e7f0b61b222d27d077bde4660c40b9 Ie57f9d0e3912632903d9740291225bfd1634ed47. Related: SYS#5315 OS#4940 OS#3787 OS#3833 Change-Id: Ie0da36124d73efc28a8809b63d7c96e2167fc412
2021-04-27 23:17:14 +00:00
conn->assignment.selected_ch_mode_rate = req->ch_mode_rate_list[i];
cosmetic scoping in reuse_existing_lchan() cosmetic prep for Ie0da36124d73efc28a8809b63d7c96e2167fc412 Related: SYS#5315 OS#4940 Change-Id: I0d1000cf108b0dfdc690f7a59d9bdf40de9043d2
2021-04-28 14:55:27 +00:00
return true;
}
assignment_fsm: Properly support assigning signalling mode TCH/x To support the 3 possible preferences, the changes needed were: - Replace 'full_rate' bool with a 3 option enum to represent the channels types for signalling - Switch from _pref/_alt to using an array sorted in preference order Originally merged as Change-Id I4c7499c8c866ea3ff7b1327edb3615d003d927d3, reverted because the change broke voice calls. Re-submitting with the fix: don't forget to set conn->assignment.requires_voice_stream. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I7513d2cbe8b695ba6f031ad11560c63a6535cf2d
2019-03-07 15:32:02 +00:00
lchan_fsm: make rsl mode-modify working again osmo-nitb supports the modification of an lchan if the lchan is compatible but in the wrong mode. This feature was dropped in the transition to AoIP/bsc-split. However, osmo-bsc still has code to generate and parse the messeages, but the FSMs do not support a mode modify yetm Lets add handling for mode-modify to the lchan_fsm and assignment_fsm in order to support mode modify again Change-Id: I2c5a283b1ee33745cc1fcfcc09a0f9382224e2eb Related: OS#4549
2020-08-21 18:47:49 +00:00
return false;
assignment_fsm: fix channel allocator preferences When the MSC allocates a channel through the ASSIGNMENT REQUEST, it may ask for a TCH/H and a TCH/F at the same time and tell which of the two types it prefers. The process of channel allocation currently selects, based on the BTS, MSC and MS capabilites exactly one apropriate codec/rate (e.g. TCH/H) and then tries to allocate it. If that allocation fails, there is no way to try the second choice and the assignment fails. For example: The MSC asks for TCH/F and TCH/H, prefering TCH/F, then the channel allocator will try TCH/F and if it fails (all TCH/F are currently in use), then TCH/H is never tried. Since the BSC currently only trys the first best codec/rate that is supported it also ignores the preference. Lets fix those problems by including the preference information and both possible codec/rate settings into the channel allocation decision. Change-Id: I5239e05c1cfbcb8af28f43373a58fa6c2d216c51 Related: OS#3503
2019-01-22 10:29:06 +00:00
}
assignment_fsm: allow assignment to a specific lchan So far the assignment FSM always tried to satisfy the channel mode and rate by either re-using the current lchan or finding a new, unused lchan. For VAMOS however, we want to pick one specific lchan. Add target_lchan to struct assignment_request and skip all mode matching and lchan selection when a specific target_lchan is set. Related: SYS#5315 OS#4940 Change-Id: I71e0d4ff4746706e0be5266e4574d70ca432e3d7
2021-04-27 22:37:14 +00:00
static int _reassignment_request(enum assign_for assign_for, struct gsm_lchan *lchan, struct gsm_lchan *to_lchan,
allow explixit TSC Set and TSC on chan activ / modif / assignment Activating / modifying to a VAMOS mode will require picking specific TSC Set / TSC. It is a bad idea to pick the TSC in each message encoding function, rather make this choice centrally. So far we pick the training sequence code to use based on the timeslot configuration, and this TSC is determined only upon encoding the RSL messages. Instead, pick the TSC to use upon the initial lchan activation / modification request; store this in the request structs and pass through the activation / modification code paths. For VAMOS modes, we also need to pick a TSC Set. Do so also upon activ / modif request. Note that the TSC Set is not yet applied in this patch, it will be applied in upcoming VAMOS patches. The activ / modif request may pass -1 for tsc_set and/or tsc to indicate no specific choice of TSC Set and TSC, resulting in the same behavior as before this patch. For example, lchan->activate.info.tsc* may be passed as -1. The exact choice for tsc_set and tsc is then stored in lchan->activate.tsc*, i.e. one level up (the .info sub-struct is considered as immutable input args). The lchan->activate.tsc* are the values actually encoded in RSL messages. After the ACK, the lchan->activate.tsc* is stored in lchan->tsc* to indicate the TSC actually in use. Same for modif. Note that in 3GPP TS 45.002, the TSC Set are numbered 1 to 4, while the TSC are numbered 0 to 7. On the wire, though, TSC Set is sent as 0 to 3 value. This is a weird discrepancy, odd choice made by the spec authors. For conformance with the spec wording, I decided to pass the TSC Set number as a 1-4 ranged value, and only convert it to the 0-3 on-the-wire format upon message encoding. So log messages and VTY output will indicate the first TSC Set as "1", but the first TSC as "0", as defined in 3GPP TS 45.002, even if that is somewhat weird. Related: SYS#5315 OS#4940 Change-Id: Ic665125255d7354f5499d10dda1dd866ab243d24
2021-05-22 14:29:35 +00:00
enum gsm_chan_t new_lchan_type, int tsc_set, int tsc)
assignment_fsm: allow assignment to a specific lchan So far the assignment FSM always tried to satisfy the channel mode and rate by either re-using the current lchan or finding a new, unused lchan. For VAMOS however, we want to pick one specific lchan. Add target_lchan to struct assignment_request and skip all mode matching and lchan selection when a specific target_lchan is set. Related: SYS#5315 OS#4940 Change-Id: I71e0d4ff4746706e0be5266e4574d70ca432e3d7
2021-04-27 22:37:14 +00:00
{
struct gsm_subscriber_connection *conn = lchan->conn;
struct assignment_request req = {
.assign_for = assign_for,
.aoip = gscon_is_aoip(conn),
.msc_assigned_cic = conn->user_plane.msc_assigned_cic,
.msc_rtp_port = conn->user_plane.msc_assigned_rtp_port,
.n_ch_mode_rate = 1,
.ch_mode_rate_list = { lchan->current_ch_mode_rate },
.target_lchan = to_lchan,
fix TSC / TSC Set used for Handover From the nature of the lchan_activate_info.tsc_set and .tsc, it is easy to forget to set tsc_set,tsc = -1 to use default TSC Set and TSC values. Handover code is one such instance that forgets to set -1. Change the semantics of tsc_set and tsc so that this kind of error can not happen again as easily: use a separate bool to flag whether to use the default config or explicit values. Implicitly fix the lchan_activate_infos "launched" in handover_fsm.c as well as abis_rsl_chan_rqd_queue_poll(). Related: OS#5244 SYS#4895 Related: I1ed6f068c85b01e5a2d7b5f2651498a1521f89af (osmo-ttcn3-hacks) Change-Id: Iae20df4387c3d75752301bd5daeeea7508966393
2021-10-02 10:03:19 +00:00
.tsc_set = {
.present = (tsc_set >= 0),
.val = tsc_set,
},
.tsc = {
.present = (tsc >= 0),
.val = tsc,
},
assignment_fsm: allow assignment to a specific lchan So far the assignment FSM always tried to satisfy the channel mode and rate by either re-using the current lchan or finding a new, unused lchan. For VAMOS however, we want to pick one specific lchan. Add target_lchan to struct assignment_request and skip all mode matching and lchan selection when a specific target_lchan is set. Related: SYS#5315 OS#4940 Change-Id: I71e0d4ff4746706e0be5266e4574d70ca432e3d7
2021-04-27 22:37:14 +00:00
};
if (to_lchan)
new_lchan_type = to_lchan->type;
req.ch_mode_rate_list[0].chan_rate = chan_t_to_chan_rate(new_lchan_type);
/* lchan activation will automatically convert chan_mode to a VAMOS equivalent if required.
* So rather always pass the plain non-VAMOS mode. */
req.ch_mode_rate_list[0].chan_mode = gsm48_chan_mode_to_non_vamos(lchan->current_ch_mode_rate.chan_mode);
OSMO_STRLCPY_ARRAY(req.msc_rtp_addr, conn->user_plane.msc_assigned_rtp_addr);
if (conn->user_plane.mgw_endpoint_ci_msc) {
req.use_osmux = osmo_mgcpc_ep_ci_get_crcx_info_to_osmux_cid(conn->user_plane.mgw_endpoint_ci_msc,
&req.osmux_cid);
}
return osmo_fsm_inst_dispatch(conn->fi, GSCON_EV_ASSIGNMENT_START, &req);
}
allow explixit TSC Set and TSC on chan activ / modif / assignment Activating / modifying to a VAMOS mode will require picking specific TSC Set / TSC. It is a bad idea to pick the TSC in each message encoding function, rather make this choice centrally. So far we pick the training sequence code to use based on the timeslot configuration, and this TSC is determined only upon encoding the RSL messages. Instead, pick the TSC to use upon the initial lchan activation / modification request; store this in the request structs and pass through the activation / modification code paths. For VAMOS modes, we also need to pick a TSC Set. Do so also upon activ / modif request. Note that the TSC Set is not yet applied in this patch, it will be applied in upcoming VAMOS patches. The activ / modif request may pass -1 for tsc_set and/or tsc to indicate no specific choice of TSC Set and TSC, resulting in the same behavior as before this patch. For example, lchan->activate.info.tsc* may be passed as -1. The exact choice for tsc_set and tsc is then stored in lchan->activate.tsc*, i.e. one level up (the .info sub-struct is considered as immutable input args). The lchan->activate.tsc* are the values actually encoded in RSL messages. After the ACK, the lchan->activate.tsc* is stored in lchan->tsc* to indicate the TSC actually in use. Same for modif. Note that in 3GPP TS 45.002, the TSC Set are numbered 1 to 4, while the TSC are numbered 0 to 7. On the wire, though, TSC Set is sent as 0 to 3 value. This is a weird discrepancy, odd choice made by the spec authors. For conformance with the spec wording, I decided to pass the TSC Set number as a 1-4 ranged value, and only convert it to the 0-3 on-the-wire format upon message encoding. So log messages and VTY output will indicate the first TSC Set as "1", but the first TSC as "0", as defined in 3GPP TS 45.002, even if that is somewhat weird. Related: SYS#5315 OS#4940 Change-Id: Ic665125255d7354f5499d10dda1dd866ab243d24
2021-05-22 14:29:35 +00:00
int reassignment_request_to_lchan(enum assign_for assign_for, struct gsm_lchan *lchan, struct gsm_lchan *to_lchan,
int tsc_set, int tsc)
assignment_fsm: allow assignment to a specific lchan So far the assignment FSM always tried to satisfy the channel mode and rate by either re-using the current lchan or finding a new, unused lchan. For VAMOS however, we want to pick one specific lchan. Add target_lchan to struct assignment_request and skip all mode matching and lchan selection when a specific target_lchan is set. Related: SYS#5315 OS#4940 Change-Id: I71e0d4ff4746706e0be5266e4574d70ca432e3d7
2021-04-27 22:37:14 +00:00
{
allow explixit TSC Set and TSC on chan activ / modif / assignment Activating / modifying to a VAMOS mode will require picking specific TSC Set / TSC. It is a bad idea to pick the TSC in each message encoding function, rather make this choice centrally. So far we pick the training sequence code to use based on the timeslot configuration, and this TSC is determined only upon encoding the RSL messages. Instead, pick the TSC to use upon the initial lchan activation / modification request; store this in the request structs and pass through the activation / modification code paths. For VAMOS modes, we also need to pick a TSC Set. Do so also upon activ / modif request. Note that the TSC Set is not yet applied in this patch, it will be applied in upcoming VAMOS patches. The activ / modif request may pass -1 for tsc_set and/or tsc to indicate no specific choice of TSC Set and TSC, resulting in the same behavior as before this patch. For example, lchan->activate.info.tsc* may be passed as -1. The exact choice for tsc_set and tsc is then stored in lchan->activate.tsc*, i.e. one level up (the .info sub-struct is considered as immutable input args). The lchan->activate.tsc* are the values actually encoded in RSL messages. After the ACK, the lchan->activate.tsc* is stored in lchan->tsc* to indicate the TSC actually in use. Same for modif. Note that in 3GPP TS 45.002, the TSC Set are numbered 1 to 4, while the TSC are numbered 0 to 7. On the wire, though, TSC Set is sent as 0 to 3 value. This is a weird discrepancy, odd choice made by the spec authors. For conformance with the spec wording, I decided to pass the TSC Set number as a 1-4 ranged value, and only convert it to the 0-3 on-the-wire format upon message encoding. So log messages and VTY output will indicate the first TSC Set as "1", but the first TSC as "0", as defined in 3GPP TS 45.002, even if that is somewhat weird. Related: SYS#5315 OS#4940 Change-Id: Ic665125255d7354f5499d10dda1dd866ab243d24
2021-05-22 14:29:35 +00:00
return _reassignment_request(assign_for, lchan, to_lchan, 0, tsc_set, tsc);
assignment_fsm: allow assignment to a specific lchan So far the assignment FSM always tried to satisfy the channel mode and rate by either re-using the current lchan or finding a new, unused lchan. For VAMOS however, we want to pick one specific lchan. Add target_lchan to struct assignment_request and skip all mode matching and lchan selection when a specific target_lchan is set. Related: SYS#5315 OS#4940 Change-Id: I71e0d4ff4746706e0be5266e4574d70ca432e3d7
2021-04-27 22:37:14 +00:00
}
int reassignment_request_to_chan_type(enum assign_for assign_for, struct gsm_lchan *lchan,
enum gsm_chan_t new_lchan_type)
{
allow explixit TSC Set and TSC on chan activ / modif / assignment Activating / modifying to a VAMOS mode will require picking specific TSC Set / TSC. It is a bad idea to pick the TSC in each message encoding function, rather make this choice centrally. So far we pick the training sequence code to use based on the timeslot configuration, and this TSC is determined only upon encoding the RSL messages. Instead, pick the TSC to use upon the initial lchan activation / modification request; store this in the request structs and pass through the activation / modification code paths. For VAMOS modes, we also need to pick a TSC Set. Do so also upon activ / modif request. Note that the TSC Set is not yet applied in this patch, it will be applied in upcoming VAMOS patches. The activ / modif request may pass -1 for tsc_set and/or tsc to indicate no specific choice of TSC Set and TSC, resulting in the same behavior as before this patch. For example, lchan->activate.info.tsc* may be passed as -1. The exact choice for tsc_set and tsc is then stored in lchan->activate.tsc*, i.e. one level up (the .info sub-struct is considered as immutable input args). The lchan->activate.tsc* are the values actually encoded in RSL messages. After the ACK, the lchan->activate.tsc* is stored in lchan->tsc* to indicate the TSC actually in use. Same for modif. Note that in 3GPP TS 45.002, the TSC Set are numbered 1 to 4, while the TSC are numbered 0 to 7. On the wire, though, TSC Set is sent as 0 to 3 value. This is a weird discrepancy, odd choice made by the spec authors. For conformance with the spec wording, I decided to pass the TSC Set number as a 1-4 ranged value, and only convert it to the 0-3 on-the-wire format upon message encoding. So log messages and VTY output will indicate the first TSC Set as "1", but the first TSC as "0", as defined in 3GPP TS 45.002, even if that is somewhat weird. Related: SYS#5315 OS#4940 Change-Id: Ic665125255d7354f5499d10dda1dd866ab243d24
2021-05-22 14:29:35 +00:00
return _reassignment_request(assign_for, lchan, NULL, new_lchan_type, -1, -1);
assignment_fsm: allow assignment to a specific lchan So far the assignment FSM always tried to satisfy the channel mode and rate by either re-using the current lchan or finding a new, unused lchan. For VAMOS however, we want to pick one specific lchan. Add target_lchan to struct assignment_request and skip all mode matching and lchan selection when a specific target_lchan is set. Related: SYS#5315 OS#4940 Change-Id: I71e0d4ff4746706e0be5266e4574d70ca432e3d7
2021-04-27 22:37:14 +00:00
}
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
void assignment_fsm_start(struct gsm_subscriber_connection *conn, struct gsm_bts *bts,
struct assignment_request *req)
{
assignment_fsm: Properly support assigning signalling mode TCH/x To support the 3 possible preferences, the changes needed were: - Replace 'full_rate' bool with a 3 option enum to represent the channels types for signalling - Switch from _pref/_alt to using an array sorted in preference order Originally merged as Change-Id I4c7499c8c866ea3ff7b1327edb3615d003d927d3, reverted because the change broke voice calls. Re-submitting with the fix: don't forget to set conn->assignment.requires_voice_stream. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I7513d2cbe8b695ba6f031ad11560c63a6535cf2d
2019-03-07 15:32:02 +00:00
static const char *rate_names[] = {
[CH_RATE_SDCCH] = "SDCCH",
[CH_RATE_HALF] = "HR",
[CH_RATE_FULL] = "FR",
};
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
struct osmo_fsm_inst *fi;
assignment_fsm: Properly support assigning signalling mode TCH/x To support the 3 possible preferences, the changes needed were: - Replace 'full_rate' bool with a 3 option enum to represent the channels types for signalling - Switch from _pref/_alt to using an array sorted in preference order Originally merged as Change-Id I4c7499c8c866ea3ff7b1327edb3615d003d927d3, reverted because the change broke voice calls. Re-submitting with the fix: don't forget to set conn->assignment.requires_voice_stream. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I7513d2cbe8b695ba6f031ad11560c63a6535cf2d
2019-03-07 15:32:02 +00:00
int i;
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
OSMO_ASSERT(conn);
OSMO_ASSERT(conn->fi);
OSMO_ASSERT(!conn->assignment.fi);
OSMO_ASSERT(!conn->assignment.new_lchan);
fi = osmo_fsm_inst_alloc_child(&assignment_fsm, conn->fi, GSCON_EV_ASSIGNMENT_END);
OSMO_ASSERT(fi);
conn->assignment.fi = fi;
fi->priv = conn;
assignment_fsm: fix channel allocator preferences When the MSC allocates a channel through the ASSIGNMENT REQUEST, it may ask for a TCH/H and a TCH/F at the same time and tell which of the two types it prefers. The process of channel allocation currently selects, based on the BTS, MSC and MS capabilites exactly one apropriate codec/rate (e.g. TCH/H) and then tries to allocate it. If that allocation fails, there is no way to try the second choice and the assignment fails. For example: The MSC asks for TCH/F and TCH/H, prefering TCH/F, then the channel allocator will try TCH/F and if it fails (all TCH/F are currently in use), then TCH/H is never tried. Since the BSC currently only trys the first best codec/rate that is supported it also ignores the preference. Lets fix those problems by including the preference information and both possible codec/rate settings into the channel allocation decision. Change-Id: I5239e05c1cfbcb8af28f43373a58fa6c2d216c51 Related: OS#3503
2019-01-22 10:29:06 +00:00
/* Create a copy of the request data and use that copy from now on. */
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
conn->assignment.req = *req;
assignment_fsm: fix channel allocator preferences When the MSC allocates a channel through the ASSIGNMENT REQUEST, it may ask for a TCH/H and a TCH/F at the same time and tell which of the two types it prefers. The process of channel allocation currently selects, based on the BTS, MSC and MS capabilites exactly one apropriate codec/rate (e.g. TCH/H) and then tries to allocate it. If that allocation fails, there is no way to try the second choice and the assignment fails. For example: The MSC asks for TCH/F and TCH/H, prefering TCH/F, then the channel allocator will try TCH/F and if it fails (all TCH/F are currently in use), then TCH/H is never tried. Since the BSC currently only trys the first best codec/rate that is supported it also ignores the preference. Lets fix those problems by including the preference information and both possible codec/rate settings into the channel allocation decision. Change-Id: I5239e05c1cfbcb8af28f43373a58fa6c2d216c51 Related: OS#3503
2019-01-22 10:29:06 +00:00
req = &conn->assignment.req;
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
remove special case from assignment_count_result() The single place that calls this special case is now actually doing the exact same thing as the general case, and this is merely code dup. Related: SYS#5315 OS#4940 Change-Id: Ic62e8a48d3c88a1966086240a41732d169328491
2021-04-28 15:21:04 +00:00
assignment_count(CTR_ASSIGNMENT_ATTEMPTED);
Fix some typos Fix typos and common misspellings in code comments and in the manual. Change-Id: I46fc9d424620c77ae9ccf78b58081bd303386d7c
2019-11-13 21:10:41 +00:00
/* Check if we need a voice stream. If yes, set the appropriate struct
assignment_fsm: fix channel allocator preferences When the MSC allocates a channel through the ASSIGNMENT REQUEST, it may ask for a TCH/H and a TCH/F at the same time and tell which of the two types it prefers. The process of channel allocation currently selects, based on the BTS, MSC and MS capabilites exactly one apropriate codec/rate (e.g. TCH/H) and then tries to allocate it. If that allocation fails, there is no way to try the second choice and the assignment fails. For example: The MSC asks for TCH/F and TCH/H, prefering TCH/F, then the channel allocator will try TCH/F and if it fails (all TCH/F are currently in use), then TCH/H is never tried. Since the BSC currently only trys the first best codec/rate that is supported it also ignores the preference. Lets fix those problems by including the preference information and both possible codec/rate settings into the channel allocation decision. Change-Id: I5239e05c1cfbcb8af28f43373a58fa6c2d216c51 Related: OS#3503
2019-01-22 10:29:06 +00:00
* members in conn */
if (check_requires_voice_stream(conn) < 0)
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
return;
assignment_fsm: allow assignment to a specific lchan So far the assignment FSM always tried to satisfy the channel mode and rate by either re-using the current lchan or finding a new, unused lchan. For VAMOS however, we want to pick one specific lchan. Add target_lchan to struct assignment_request and skip all mode matching and lchan selection when a specific target_lchan is set. Related: SYS#5315 OS#4940 Change-Id: I71e0d4ff4746706e0be5266e4574d70ca432e3d7
2021-04-27 22:37:14 +00:00
if (!req->target_lchan && reuse_existing_lchan(conn)) {
/* The already existing lchan is suitable for this mode */
lchan and assignment FSMs: make Channel Mode Modify more sane The Channel Mode Modify procedure is currently implemented for changing a TCH lchan from signalling to voice mode. For that, however, it is re-using (abusing) the channel activation structs and state transitions, and thus always implies activating a voice stream when the mode modification is done. I will add a Channel Mode Modify to enable VAMOS mode soon, so I require separate structs and state transitions which also work on an lchan that already has a voice stream established: a struct lchan_modify_info and LCHAN_EV_REQUEST_MODE_MODIFY, and dedicated assignment FSM state ASSIGNMENT_ST_WAIT_LCHAN_MODIFIED. For the part where a Channel Mode Modify enables a voice stream after switching from signalling to speech mode, still use the channel activation code path, but only once the mode modification is done. General improvements: - To ask for a mode modification, emit an FSM event that ensures a mode modify only happens when the lchan state allows it. - The new lchan_modify_info struct reflects only those parts that have an effect during a mode modification (before the lchan_activate_info was fully populated, many values not having an effect). - More accurate logging, indicating "Mode Modify" instead of "Channel Activation" A TTCN3 test for the Channel Mode Modify procedure is added in Idf4efaed986de0bbd2b663313e837352cc139f0f, and the test passes both before and after this patch is applied. Related: SYS#4895 Change-Id: I4986844f839b1c9672c61d916eb3d33d0042d747
2021-04-14 17:39:01 +00:00
conn->assignment.new_lchan = NULL;
lchan_fsm: make rsl mode-modify working again osmo-nitb supports the modification of an lchan if the lchan is compatible but in the wrong mode. This feature was dropped in the transition to AoIP/bsc-split. However, osmo-bsc still has code to generate and parse the messeages, but the FSMs do not support a mode modify yetm Lets add handling for mode-modify to the lchan_fsm and assignment_fsm in order to support mode modify again Change-Id: I2c5a283b1ee33745cc1fcfcc09a0f9382224e2eb Related: OS#4549
2020-08-21 18:47:49 +00:00
/* If the requested mode and the current TCH mode matches up, just send the
* assignment complete directly and be done with the assignment procedure. */
make sure channel mode and s15_s0 are updated only after an ACK I noticed during testing that an lchan used as TCH/F in fact still had its channel mode set to Signalling -- because on Assignment, the Speech mode used to be placed in the *previous* lchan and the new lchan was never updated after the Activ ACK. This is unbearable confusion which I complained about numerous times, so far mostly for cosmetic reasons. But implementing re-assignment properly actually requires this to be cleaned up. Keep all volatile chan mode settings in lchan->activate.* or lchan->modify.*, and only update lchan->* members when an ACK has been received for those settings. So a failed request keeps a sane state. Make sure that those settings are in fact updated in the proper lchan, upon an ACK, so that subsequent re-assignment or mode-modify know the accurate lchan state. Related are upcoming patches that sort out the AMR multirate configuration in a similar fashion, see Iebac2dc26412d877e5364f90d6f2ed7a7952351e Ia7519d2fa9e7f0b61b222d27d077bde4660c40b9 Ie57f9d0e3912632903d9740291225bfd1634ed47. Related: SYS#5315 OS#4940 OS#3787 OS#3833 Change-Id: Ie0da36124d73efc28a8809b63d7c96e2167fc412
2021-04-27 23:17:14 +00:00
if (conn->lchan->current_ch_mode_rate.chan_mode == conn->assignment.selected_ch_mode_rate.chan_mode) {
lchan_fsm: make rsl mode-modify working again osmo-nitb supports the modification of an lchan if the lchan is compatible but in the wrong mode. This feature was dropped in the transition to AoIP/bsc-split. However, osmo-bsc still has code to generate and parse the messeages, but the FSMs do not support a mode modify yetm Lets add handling for mode-modify to the lchan_fsm and assignment_fsm in order to support mode modify again Change-Id: I2c5a283b1ee33745cc1fcfcc09a0f9382224e2eb Related: OS#4549
2020-08-21 18:47:49 +00:00
LOG_ASSIGNMENT(conn, LOGL_DEBUG,
"Current lchan mode is compatible with requested chan_mode,"
" sending BSSMAP Assignment Complete directly."
" requested chan_mode=%s; current lchan is %s\n",
make sure channel mode and s15_s0 are updated only after an ACK I noticed during testing that an lchan used as TCH/F in fact still had its channel mode set to Signalling -- because on Assignment, the Speech mode used to be placed in the *previous* lchan and the new lchan was never updated after the Activ ACK. This is unbearable confusion which I complained about numerous times, so far mostly for cosmetic reasons. But implementing re-assignment properly actually requires this to be cleaned up. Keep all volatile chan mode settings in lchan->activate.* or lchan->modify.*, and only update lchan->* members when an ACK has been received for those settings. So a failed request keeps a sane state. Make sure that those settings are in fact updated in the proper lchan, upon an ACK, so that subsequent re-assignment or mode-modify know the accurate lchan state. Related are upcoming patches that sort out the AMR multirate configuration in a similar fashion, see Iebac2dc26412d877e5364f90d6f2ed7a7952351e Ia7519d2fa9e7f0b61b222d27d077bde4660c40b9 Ie57f9d0e3912632903d9740291225bfd1634ed47. Related: SYS#5315 OS#4940 OS#3787 OS#3833 Change-Id: Ie0da36124d73efc28a8809b63d7c96e2167fc412
2021-04-27 23:17:14 +00:00
gsm48_chan_mode_name(conn->assignment.selected_ch_mode_rate.chan_mode),
lchan_fsm: make rsl mode-modify working again osmo-nitb supports the modification of an lchan if the lchan is compatible but in the wrong mode. This feature was dropped in the transition to AoIP/bsc-split. However, osmo-bsc still has code to generate and parse the messeages, but the FSMs do not support a mode modify yetm Lets add handling for mode-modify to the lchan_fsm and assignment_fsm in order to support mode modify again Change-Id: I2c5a283b1ee33745cc1fcfcc09a0f9382224e2eb Related: OS#4549
2020-08-21 18:47:49 +00:00
gsm_lchan_name(conn->lchan));
assignment_fsm: send BSSMAP response only after Assignment Request So far, only the MSC asked for Assignment via Assignment Request, which we answer with a BSSMAP Assignment Complete or Assignment Failure when done. When Assignment is triggered for any other reason (congestion resolution, VAMOS, VTY), we will not send any such messages to the MSC. Additional enum values will be added in subsequent commits: Id56a890106b93fcee67ac9401b890e7b63bba421 ASSIGN_FOR_CONGESTION_RESOLUTION If006f5caaf83b07675f57e5665cfa79328da55e6 ASSIGN_FOR_VTY Related: SYS#5315 OS#4940 Change-Id: Ie0cddbdb00abcec78e153f4ae6d04ce75080a111
2021-05-01 02:52:05 +00:00
if (req->assign_for == ASSIGN_FOR_BSSMAP_REQ)
send_assignment_complete(conn);
lchan_fsm: make rsl mode-modify working again osmo-nitb supports the modification of an lchan if the lchan is compatible but in the wrong mode. This feature was dropped in the transition to AoIP/bsc-split. However, osmo-bsc still has code to generate and parse the messeages, but the FSMs do not support a mode modify yetm Lets add handling for mode-modify to the lchan_fsm and assignment_fsm in order to support mode modify again Change-Id: I2c5a283b1ee33745cc1fcfcc09a0f9382224e2eb Related: OS#4549
2020-08-21 18:47:49 +00:00
/* If something went wrong during send_assignment_complete(),
* the fi will be gone from error handling in there. */
if (conn->assignment.fi) {
assignment_count_result(CTR_ASSIGNMENT_COMPLETED);
osmo_fsm_inst_term(conn->assignment.fi, OSMO_FSM_TERM_REGULAR, 0);
}
return;
}
/* The requested mode does not match the current TCH mode but the lchan is
* compatible. We will initiate a mode modify procedure. */
assignment_fsm: fix channel allocator preferences When the MSC allocates a channel through the ASSIGNMENT REQUEST, it may ask for a TCH/H and a TCH/F at the same time and tell which of the two types it prefers. The process of channel allocation currently selects, based on the BTS, MSC and MS capabilites exactly one apropriate codec/rate (e.g. TCH/H) and then tries to allocate it. If that allocation fails, there is no way to try the second choice and the assignment fails. For example: The MSC asks for TCH/F and TCH/H, prefering TCH/F, then the channel allocator will try TCH/F and if it fails (all TCH/F are currently in use), then TCH/H is never tried. Since the BSC currently only trys the first best codec/rate that is supported it also ignores the preference. Lets fix those problems by including the preference information and both possible codec/rate settings into the channel allocation decision. Change-Id: I5239e05c1cfbcb8af28f43373a58fa6c2d216c51 Related: OS#3503
2019-01-22 10:29:06 +00:00
LOG_ASSIGNMENT(conn, LOGL_DEBUG,
lchan_fsm: make rsl mode-modify working again osmo-nitb supports the modification of an lchan if the lchan is compatible but in the wrong mode. This feature was dropped in the transition to AoIP/bsc-split. However, osmo-bsc still has code to generate and parse the messeages, but the FSMs do not support a mode modify yetm Lets add handling for mode-modify to the lchan_fsm and assignment_fsm in order to support mode modify again Change-Id: I2c5a283b1ee33745cc1fcfcc09a0f9382224e2eb Related: OS#4549
2020-08-21 18:47:49 +00:00
"Current lchan mode is not compatible with requested chan_mode,"
" so we will modify it. requested chan_mode=%s; current lchan is %s\n",
make sure channel mode and s15_s0 are updated only after an ACK I noticed during testing that an lchan used as TCH/F in fact still had its channel mode set to Signalling -- because on Assignment, the Speech mode used to be placed in the *previous* lchan and the new lchan was never updated after the Activ ACK. This is unbearable confusion which I complained about numerous times, so far mostly for cosmetic reasons. But implementing re-assignment properly actually requires this to be cleaned up. Keep all volatile chan mode settings in lchan->activate.* or lchan->modify.*, and only update lchan->* members when an ACK has been received for those settings. So a failed request keeps a sane state. Make sure that those settings are in fact updated in the proper lchan, upon an ACK, so that subsequent re-assignment or mode-modify know the accurate lchan state. Related are upcoming patches that sort out the AMR multirate configuration in a similar fashion, see Iebac2dc26412d877e5364f90d6f2ed7a7952351e Ia7519d2fa9e7f0b61b222d27d077bde4660c40b9 Ie57f9d0e3912632903d9740291225bfd1634ed47. Related: SYS#5315 OS#4940 OS#3787 OS#3833 Change-Id: Ie0da36124d73efc28a8809b63d7c96e2167fc412
2021-04-27 23:17:14 +00:00
gsm48_chan_mode_name(conn->assignment.selected_ch_mode_rate.chan_mode),
assignment_fsm: fix channel allocator preferences When the MSC allocates a channel through the ASSIGNMENT REQUEST, it may ask for a TCH/H and a TCH/F at the same time and tell which of the two types it prefers. The process of channel allocation currently selects, based on the BTS, MSC and MS capabilites exactly one apropriate codec/rate (e.g. TCH/H) and then tries to allocate it. If that allocation fails, there is no way to try the second choice and the assignment fails. For example: The MSC asks for TCH/F and TCH/H, prefering TCH/F, then the channel allocator will try TCH/F and if it fails (all TCH/F are currently in use), then TCH/H is never tried. Since the BSC currently only trys the first best codec/rate that is supported it also ignores the preference. Lets fix those problems by including the preference information and both possible codec/rate settings into the channel allocation decision. Change-Id: I5239e05c1cfbcb8af28f43373a58fa6c2d216c51 Related: OS#3503
2019-01-22 10:29:06 +00:00
gsm_lchan_name(conn->lchan));
assignment_fsm: tweak state transitions (prep for reassignment) It is better design to take state change actions in the onenter function, instead of triggering a state change and then calling lchan_mode_modify() or lchan_activate(). The reason is that in principle, any state change may cause error handling to abort and deallocate FSMs. This is also preparation for reassignment to a specific lchan in an upcoming patch. Related: SYS#5315 OS#4940 OS#3277 Change-Id: I9a2e7eefd4196b80671311e5dfd275893ec0e275
2021-05-18 22:56:07 +00:00
assignment_fsm_state_chg(ASSIGNMENT_ST_WAIT_LCHAN_MODIFIED);
assignment_fsm: fix channel allocator preferences When the MSC allocates a channel through the ASSIGNMENT REQUEST, it may ask for a TCH/H and a TCH/F at the same time and tell which of the two types it prefers. The process of channel allocation currently selects, based on the BTS, MSC and MS capabilites exactly one apropriate codec/rate (e.g. TCH/H) and then tries to allocate it. If that allocation fails, there is no way to try the second choice and the assignment fails. For example: The MSC asks for TCH/F and TCH/H, prefering TCH/F, then the channel allocator will try TCH/F and if it fails (all TCH/F are currently in use), then TCH/H is never tried. Since the BSC currently only trys the first best codec/rate that is supported it also ignores the preference. Lets fix those problems by including the preference information and both possible codec/rate settings into the channel allocation decision. Change-Id: I5239e05c1cfbcb8af28f43373a58fa6c2d216c51 Related: OS#3503
2019-01-22 10:29:06 +00:00
return;
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
}
assignment_fsm: allow assignment to a specific lchan So far the assignment FSM always tried to satisfy the channel mode and rate by either re-using the current lchan or finding a new, unused lchan. For VAMOS however, we want to pick one specific lchan. Add target_lchan to struct assignment_request and skip all mode matching and lchan selection when a specific target_lchan is set. Related: SYS#5315 OS#4940 Change-Id: I71e0d4ff4746706e0be5266e4574d70ca432e3d7
2021-04-27 22:37:14 +00:00
if (req->target_lchan) {
bool matching_mode;
make sure channel mode and s15_s0 are updated only after an ACK I noticed during testing that an lchan used as TCH/F in fact still had its channel mode set to Signalling -- because on Assignment, the Speech mode used to be placed in the *previous* lchan and the new lchan was never updated after the Activ ACK. This is unbearable confusion which I complained about numerous times, so far mostly for cosmetic reasons. But implementing re-assignment properly actually requires this to be cleaned up. Keep all volatile chan mode settings in lchan->activate.* or lchan->modify.*, and only update lchan->* members when an ACK has been received for those settings. So a failed request keeps a sane state. Make sure that those settings are in fact updated in the proper lchan, upon an ACK, so that subsequent re-assignment or mode-modify know the accurate lchan state. Related are upcoming patches that sort out the AMR multirate configuration in a similar fashion, see Iebac2dc26412d877e5364f90d6f2ed7a7952351e Ia7519d2fa9e7f0b61b222d27d077bde4660c40b9 Ie57f9d0e3912632903d9740291225bfd1634ed47. Related: SYS#5315 OS#4940 OS#3787 OS#3833 Change-Id: Ie0da36124d73efc28a8809b63d7c96e2167fc412
2021-04-27 23:17:14 +00:00
assignment_fsm: allow assignment to a specific lchan So far the assignment FSM always tried to satisfy the channel mode and rate by either re-using the current lchan or finding a new, unused lchan. For VAMOS however, we want to pick one specific lchan. Add target_lchan to struct assignment_request and skip all mode matching and lchan selection when a specific target_lchan is set. Related: SYS#5315 OS#4940 Change-Id: I71e0d4ff4746706e0be5266e4574d70ca432e3d7
2021-04-27 22:37:14 +00:00
/* The caller already picked a target lchan to assign to. No need to try re-using the current lchan or
* picking a new one. */
if (!lchan_state_is(req->target_lchan, LCHAN_ST_UNUSED)) {
assignment_fail(GSM0808_CAUSE_NO_RADIO_RESOURCE_AVAILABLE,
"Assignment to lchan %s requested, but lchan is already in use (state=%s)\n",
gsm_lchan_name(req->target_lchan),
osmo_fsm_inst_state_name(req->target_lchan->fi));
return;
}
conn->assignment.new_lchan = req->target_lchan;
matching_mode = false;
for (i = 0; i < req->n_ch_mode_rate; i++) {
if (!lchan_type_compat_with_mode(conn->assignment.new_lchan->type, &req->ch_mode_rate_list[i]))
continue;
conn->assignment.selected_ch_mode_rate = req->ch_mode_rate_list[i];
matching_mode = true;
}
if (!matching_mode) {
assignment_fsm: Add assert to guard ptr access Let's make sure the null pointer is caught by the assert if ever code ends up here with conn->lchan being NULL. Change-Id: I404df638da6a93caa535f10fd330ea24a775bfc3
2021-07-05 15:20:18 +00:00
OSMO_ASSERT(conn->lchan);
assignment_fsm: allow assignment to a specific lchan So far the assignment FSM always tried to satisfy the channel mode and rate by either re-using the current lchan or finding a new, unused lchan. For VAMOS however, we want to pick one specific lchan. Add target_lchan to struct assignment_request and skip all mode matching and lchan selection when a specific target_lchan is set. Related: SYS#5315 OS#4940 Change-Id: I71e0d4ff4746706e0be5266e4574d70ca432e3d7
2021-04-27 22:37:14 +00:00
assignment_fail(GSM0808_CAUSE_NO_RADIO_RESOURCE_AVAILABLE,
log: assignment_fsm: tweak err msg for incompatible chan Related: SYS#5315 OS#4940 Change-Id: Ifd8790923d54a27061708ff8077d509238b9c7dd
2021-05-19 00:41:28 +00:00
"Assignment of lchan %s to %s type %s requested, but lchan is not compatible",
gsm_lchan_name(conn->lchan),
gsm_lchan_name(req->target_lchan),
gsm_lchant_name(conn->assignment.new_lchan->type));
assignment_fsm: allow assignment to a specific lchan So far the assignment FSM always tried to satisfy the channel mode and rate by either re-using the current lchan or finding a new, unused lchan. For VAMOS however, we want to pick one specific lchan. Add target_lchan to struct assignment_request and skip all mode matching and lchan selection when a specific target_lchan is set. Related: SYS#5315 OS#4940 Change-Id: I71e0d4ff4746706e0be5266e4574d70ca432e3d7
2021-04-27 22:37:14 +00:00
return;
}
} else {
/* Try to allocate a new lchan in order of preference */
for (i = 0; i < req->n_ch_mode_rate; i++) {
conn->assignment.new_lchan = lchan_select_by_chan_mode(bts,
lchan_select: allow different alloc order for assignment and handover A follow-up patch implements a special channel allocation mode, which is only working for assignment (basically TCH selection for a voice call). This mode cannot be employed for initial CHANNEL REQUEST or handover due to the absence of an already established lchan. Adding this mode to the existing VTY command syntax would be confusing: channel allocator (ascending|desscending|dynamic) ^^^^^^^ so this patch extends the VTY syntax in a way that it becomes possible to configure different channel allocator modes for different cases: OsmoBSC(config-net-bts)# channel allocator mode ? set-all Set a single mode for all variants chan-req Channel allocation for CHANNEL REQUEST (RACH) assignment Channel allocation for assignment handover Channel allocation for handover The old command syntax, which is basically 'set-all', is kept for backwards compatibility, but marked as deprecated. Change-Id: I3ae73b36ee9433cc768376b56f0765e5f416162f Related: SYS#5460
2022-06-05 19:43:28 +00:00
req->ch_mode_rate_list[i].chan_mode,
req->ch_mode_rate_list[i].chan_rate,
lchan_select: implement dynamic selection mode for assignment This change implements an additional channel allocation mode, which can be employed during a TCH channel allocation for assignment. Selection between ascending and descending order is performed depending on pre-configured parameters: * Uplink RxLev threshold and number of samples for averaging, * C0 (BCCH carrier) channel load threshold. This is useful in setups where Tx power of the RF carriers cannot be adjusted +dynamically at run-time and thus BS Power Control cannot be performed. In such setups the BCCH carrier is transmitting at relatively higher power than the other RF carriers. The key idea is to allocate channels in a smarter way, so that UEs with poor signal would get channels on carriers with high Tx power, while UEs with good signal could use carriers with lower Tx power. Change-Id: I1b7a0d706976b73cc5c30a8714b830811addfe8d Related: osmo-ttcn3-hacks.git Ia522f37c1c001b3a36f5145b8875fbb88311c2e5 Related: SYS#5460
2022-06-15 19:10:16 +00:00
SELECT_FOR_ASSIGNMENT, conn->lchan);
assignment_fsm: allow assignment to a specific lchan So far the assignment FSM always tried to satisfy the channel mode and rate by either re-using the current lchan or finding a new, unused lchan. For VAMOS however, we want to pick one specific lchan. Add target_lchan to struct assignment_request and skip all mode matching and lchan selection when a specific target_lchan is set. Related: SYS#5315 OS#4940 Change-Id: I71e0d4ff4746706e0be5266e4574d70ca432e3d7
2021-04-27 22:37:14 +00:00
if (!conn->assignment.new_lchan)
continue;
LOG_ASSIGNMENT(conn, LOGL_DEBUG, "selected new lchan %s for mode[%d] = %s channel_rate=%d\n",
gsm_lchan_name(conn->assignment.new_lchan),
i, gsm48_chan_mode_name(req->ch_mode_rate_list[i].chan_mode),
req->ch_mode_rate_list[i].chan_rate);
conn->assignment.selected_ch_mode_rate = req->ch_mode_rate_list[i];
break;
}
assignment_fsm: fix channel allocator preferences When the MSC allocates a channel through the ASSIGNMENT REQUEST, it may ask for a TCH/H and a TCH/F at the same time and tell which of the two types it prefers. The process of channel allocation currently selects, based on the BTS, MSC and MS capabilites exactly one apropriate codec/rate (e.g. TCH/H) and then tries to allocate it. If that allocation fails, there is no way to try the second choice and the assignment fails. For example: The MSC asks for TCH/F and TCH/H, prefering TCH/F, then the channel allocator will try TCH/F and if it fails (all TCH/F are currently in use), then TCH/H is never tried. Since the BSC currently only trys the first best codec/rate that is supported it also ignores the preference. Lets fix those problems by including the preference information and both possible codec/rate settings into the channel allocation decision. Change-Id: I5239e05c1cfbcb8af28f43373a58fa6c2d216c51 Related: OS#3503
2019-01-22 10:29:06 +00:00
}
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
assignment_fsm: fix channel allocator preferences When the MSC allocates a channel through the ASSIGNMENT REQUEST, it may ask for a TCH/H and a TCH/F at the same time and tell which of the two types it prefers. The process of channel allocation currently selects, based on the BTS, MSC and MS capabilites exactly one apropriate codec/rate (e.g. TCH/H) and then tries to allocate it. If that allocation fails, there is no way to try the second choice and the assignment fails. For example: The MSC asks for TCH/F and TCH/H, prefering TCH/F, then the channel allocator will try TCH/F and if it fails (all TCH/F are currently in use), then TCH/H is never tried. Since the BSC currently only trys the first best codec/rate that is supported it also ignores the preference. Lets fix those problems by including the preference information and both possible codec/rate settings into the channel allocation decision. Change-Id: I5239e05c1cfbcb8af28f43373a58fa6c2d216c51 Related: OS#3503
2019-01-22 10:29:06 +00:00
/* Check whether the lchan allocation was successful or not and tear
* down the assignment in case of failure. */
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
if (!conn->assignment.new_lchan) {
Count assignment rates per BTS as well This adds the assignment counters for the BTS as well and changes the assignment_count() macro to increase both the counters for the BSC as well as the BTS. Related: SYS#4877 Change-Id: I0009e51d4caf68e762138d98e2e23d49acc3cc1a
2020-06-08 12:36:46 +00:00
assignment_count_result(CTR_ASSIGNMENT_NO_CHANNEL);
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
assignment_fail(GSM0808_CAUSE_NO_RADIO_RESOURCE_AVAILABLE,
"BSSMAP Assignment Command:"
log: assignment_fsm: drop newline from assignment_fail Drop terminating newline from an assignment_fail() call which causes an erratic line feed in logging. Change-Id: Ib56e1f6a45c7e2c235f8ef0c8dea7151481677ab
2021-05-23 18:48:39 +00:00
" No lchan available for: pref=%s:%s / alt1=%s:%s / alt2=%s:%s",
make sure channel mode and s15_s0 are updated only after an ACK I noticed during testing that an lchan used as TCH/F in fact still had its channel mode set to Signalling -- because on Assignment, the Speech mode used to be placed in the *previous* lchan and the new lchan was never updated after the Activ ACK. This is unbearable confusion which I complained about numerous times, so far mostly for cosmetic reasons. But implementing re-assignment properly actually requires this to be cleaned up. Keep all volatile chan mode settings in lchan->activate.* or lchan->modify.*, and only update lchan->* members when an ACK has been received for those settings. So a failed request keeps a sane state. Make sure that those settings are in fact updated in the proper lchan, upon an ACK, so that subsequent re-assignment or mode-modify know the accurate lchan state. Related are upcoming patches that sort out the AMR multirate configuration in a similar fashion, see Iebac2dc26412d877e5364f90d6f2ed7a7952351e Ia7519d2fa9e7f0b61b222d27d077bde4660c40b9 Ie57f9d0e3912632903d9740291225bfd1634ed47. Related: SYS#5315 OS#4940 OS#3787 OS#3833 Change-Id: Ie0da36124d73efc28a8809b63d7c96e2167fc412
2021-04-27 23:17:14 +00:00
gsm48_chan_mode_name(req->ch_mode_rate_list[0].chan_mode),
rate_names[req->ch_mode_rate_list[0].chan_rate],
req->n_ch_mode_rate > 1 ? gsm48_chan_mode_name(req->ch_mode_rate_list[1].chan_mode) : "",
req->n_ch_mode_rate > 1 ? rate_names[req->ch_mode_rate_list[1].chan_rate] : "",
req->n_ch_mode_rate > 2 ? gsm48_chan_mode_name(req->ch_mode_rate_list[2].chan_mode) : "",
req->n_ch_mode_rate > 2 ? rate_names[req->ch_mode_rate_list[2].chan_rate] : ""
assignment_fsm: Properly support assigning signalling mode TCH/x To support the 3 possible preferences, the changes needed were: - Replace 'full_rate' bool with a 3 option enum to represent the channels types for signalling - Switch from _pref/_alt to using an array sorted in preference order Originally merged as Change-Id I4c7499c8c866ea3ff7b1327edb3615d003d927d3, reverted because the change broke voice calls. Re-submitting with the fix: don't forget to set conn->assignment.requires_voice_stream. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I7513d2cbe8b695ba6f031ad11560c63a6535cf2d
2019-03-07 15:32:02 +00:00
);
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
return;
}
assignment_fsm_update_id(conn);
assignment_fsm: Properly support assigning signalling mode TCH/x To support the 3 possible preferences, the changes needed were: - Replace 'full_rate' bool with a 3 option enum to represent the channels types for signalling - Switch from _pref/_alt to using an array sorted in preference order Originally merged as Change-Id I4c7499c8c866ea3ff7b1327edb3615d003d927d3, reverted because the change broke voice calls. Re-submitting with the fix: don't forget to set conn->assignment.requires_voice_stream. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I7513d2cbe8b695ba6f031ad11560c63a6535cf2d
2019-03-07 15:32:02 +00:00
LOG_ASSIGNMENT(conn, LOGL_INFO, "Starting Assignment: chan_mode=%s, chan_type=%s,"
bssap: Feed Assign Complete with Osmux CID retrieved from MGW The CID is sent back to the MSC so it can configure its MGW to send Osmux frames to the BSC MGW using this CID. Change-Id: I77dfdd965ae828c39a9818669177aefd22bc02f2
2019-04-16 18:33:25 +00:00
" aoip=%s MSC-rtp=%s:%u (osmux=%s)\n",
make sure channel mode and s15_s0 are updated only after an ACK I noticed during testing that an lchan used as TCH/F in fact still had its channel mode set to Signalling -- because on Assignment, the Speech mode used to be placed in the *previous* lchan and the new lchan was never updated after the Activ ACK. This is unbearable confusion which I complained about numerous times, so far mostly for cosmetic reasons. But implementing re-assignment properly actually requires this to be cleaned up. Keep all volatile chan mode settings in lchan->activate.* or lchan->modify.*, and only update lchan->* members when an ACK has been received for those settings. So a failed request keeps a sane state. Make sure that those settings are in fact updated in the proper lchan, upon an ACK, so that subsequent re-assignment or mode-modify know the accurate lchan state. Related are upcoming patches that sort out the AMR multirate configuration in a similar fashion, see Iebac2dc26412d877e5364f90d6f2ed7a7952351e Ia7519d2fa9e7f0b61b222d27d077bde4660c40b9 Ie57f9d0e3912632903d9740291225bfd1634ed47. Related: SYS#5315 OS#4940 OS#3787 OS#3833 Change-Id: Ie0da36124d73efc28a8809b63d7c96e2167fc412
2021-04-27 23:17:14 +00:00
gsm48_chan_mode_name(conn->assignment.selected_ch_mode_rate.chan_mode),
rate_names[conn->assignment.selected_ch_mode_rate.chan_rate],
bssap: Feed Assign Complete with Osmux CID retrieved from MGW The CID is sent back to the MSC so it can configure its MGW to send Osmux frames to the BSC MGW using this CID. Change-Id: I77dfdd965ae828c39a9818669177aefd22bc02f2
2019-04-16 18:33:25 +00:00
req->aoip ? "yes" : "no", req->msc_rtp_addr, req->msc_rtp_port,
req->use_osmux ? "yes" : "no");
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
assignment_fsm_state_chg(ASSIGNMENT_ST_WAIT_LCHAN_ACTIVE);
assignment_fsm: tweak state transitions (prep for reassignment) It is better design to take state change actions in the onenter function, instead of triggering a state change and then calling lchan_mode_modify() or lchan_activate(). The reason is that in principle, any state change may cause error handling to abort and deallocate FSMs. This is also preparation for reassignment to a specific lchan in an upcoming patch. Related: SYS#5315 OS#4940 OS#3277 Change-Id: I9a2e7eefd4196b80671311e5dfd275893ec0e275
2021-05-18 22:56:07 +00:00
}
static void assignment_fsm_wait_lchan_active_onenter(struct osmo_fsm_inst *fi, uint32_t prev_state)
{
struct gsm_subscriber_connection *conn = assignment_fi_conn(fi);
struct assignment_request *req = &conn->assignment.req;
struct lchan_activate_info activ_info = {
cosmetic: rename FOR_* to ACTIVATE_FOR_* Soon, there will also be enums with ASSIGNMENT_FOR_* and MODIFY_FOR_* naming. Add the ACTIVATE_ prefix to the existing enum to clarify. Change-Id: I12190d4d154a1da6a9ebc9a755ccc2fe382ff188
2021-05-10 15:47:46 +00:00
.activ_for = ACTIVATE_FOR_ASSIGNMENT,
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
.for_conn = conn,
make sure channel mode and s15_s0 are updated only after an ACK I noticed during testing that an lchan used as TCH/F in fact still had its channel mode set to Signalling -- because on Assignment, the Speech mode used to be placed in the *previous* lchan and the new lchan was never updated after the Activ ACK. This is unbearable confusion which I complained about numerous times, so far mostly for cosmetic reasons. But implementing re-assignment properly actually requires this to be cleaned up. Keep all volatile chan mode settings in lchan->activate.* or lchan->modify.*, and only update lchan->* members when an ACK has been received for those settings. So a failed request keeps a sane state. Make sure that those settings are in fact updated in the proper lchan, upon an ACK, so that subsequent re-assignment or mode-modify know the accurate lchan state. Related are upcoming patches that sort out the AMR multirate configuration in a similar fashion, see Iebac2dc26412d877e5364f90d6f2ed7a7952351e Ia7519d2fa9e7f0b61b222d27d077bde4660c40b9 Ie57f9d0e3912632903d9740291225bfd1634ed47. Related: SYS#5315 OS#4940 OS#3787 OS#3833 Change-Id: Ie0da36124d73efc28a8809b63d7c96e2167fc412
2021-04-27 23:17:14 +00:00
.ch_mode_rate = conn->assignment.selected_ch_mode_rate,
lchan activation: add explicit encryption info to activation For intra-BSC handover, the previous encryption is copied from the old lchan, which of course is not available during inter-BSC handover. Hence the lchan activation info needs to include an explicit encryption information, and we must not rely on the presence of the previous lchan to copy encryption information from. Add struct lchan_activate_info.encr to allow passing encryption info through lchan_activate() without requiring a previous struct gsm_lchan to be present. Instead of copying from the old lchan, always copy encryption info to lchan_activate_info, and during activation, just before sending the Channel Activation, copy the lchan_activate_info.encr to the new lchan. This prepares for upcoming I5b269f50bd2092516bfdf87746196983d3ac49d1 which obtains the encryption information from an intra-BSC-incoming Handover Request message. Related: OS#3842 Related: I5b269f50bd2092516bfdf87746196983d3ac49d1 Change-Id: Ib3d259a5711add65ab7298bfa3977855a17a1642
2019-03-15 01:49:18 +00:00
.encr = conn->lchan->encr,
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
.requires_voice_stream = conn->assignment.requires_voice_stream,
.msc_assigned_cic = req->msc_assigned_cic,
make sure early lchan act failure resets the lchan Fix crash after AMR configuration fails. The crash is due to an assertion that finds a non-NULL conn in the lchan, when re-using an lchan that has failed in AMR configuration earlier on. That is because the AMR config still happens in state UNUSED. DCHAN ERROR lchan(0-0-2-TCH_F_TCH_H_PDCH-0)[0x6120000066a0]{UNUSED}: (type=TCH_F) lchan allocation failed in state UNUSED: Can not generate multirate configuration IE ... DCHAN DEBUG lchan(0-0-2-TCH_F_TCH_H_PDCH-0)[0x6120000066a0]{UNUSED}: (type=TCH_F) After failure handling, already in state UNUSED ... ... DCHAN DEBUG lchan(0-0-2-TCH_F_TCH_H_PDCH-0)[0x6120000066a0]{UNUSED}: Received Event LCHAN_EV_ACTIVATE (lchan_fsm.c:324) Assert failed !lchan->conn ../../../../src/osmo-bsc/src/osmo-bsc/lchan_fsm.c:491 The FSM design idea is that when returning to the UNUSED state, all lchan state is cleared. However, when calling lchan_activate(), a failure may happen still in state UNUSED, so that we don't transition *back* to UNUSED properly. So, first transition out of UNUSED before failures can happen. (Other ways to solve this would be to invoke lchan clearing even if already in UNUSED, but semantically, transitioning first makes more sense.) Upon LCHAN_EV_ACTIVATE, just remember the lchan_activate_info and transition to WAIT_TS_READY, so that on lchan_fail(), we can normally transition back to UNUSED and clear the lchan. Move the initial lchan activation code to lchan_fsm_wait_ts_ready_onenter(). Also, there is a bit of duplication of members of the lchan->activate (lchan state) and the lchan_activate_info (passed to lchan_activate()) structs. The fix for this also removes the dup: Add struct lchan_activate_info as child struct at lchan->activate.info, drop the other lchan->activate members that would dup .info.*. Move struct lchan_activate_info declaration to gsm_data.h. Apply the new '.info' member struct throughout the code. Related: OS#3737 Change-Id: Ide665b10fa3f4583059c55346db8da833959e3cc
2018-12-19 16:19:02 +00:00
.re_use_mgw_endpoint_from_lchan = conn->lchan,
lchan activation: indicate whether TA is known On lchan activation, we already know the Timing Advance in most situations: from the Channel Request RACH, or from a previous lchan in the same cell. Place this information in lchan->activate.info.ta. So far, the lchan->last_ta (until recently called rqd_ta) was used to store the initial TA for channel activation -- move the initial TA to lchan->activate.info.ta, for proper scoping. Only an inter-cell handover does not yet know a Timing Advance (until the Handover Detection RACH is received), so indicate activate.info.ta_known = false for that case. If ta_known is false, do not include an Access Delay IE in the Channel Activation message, ensuring that the BTS does not use an arbitrary TA that is likely inaccurate. The effect for OsmoBTS is that we will *not* start the downlink SACCH on channel activation for inter-cell handover, but will wait for a HO RACH first, and then use the correct TA when enabling downlink SACCH. Related: OS#4008 SYS#5192 Change-Id: I986bf93e8acd6aef7eaf63ac962480b680aa894f
2021-01-27 22:45:33 +00:00
.ta = conn->lchan->last_ta,
.ta_known = true,
allow explixit TSC Set and TSC on chan activ / modif / assignment Activating / modifying to a VAMOS mode will require picking specific TSC Set / TSC. It is a bad idea to pick the TSC in each message encoding function, rather make this choice centrally. So far we pick the training sequence code to use based on the timeslot configuration, and this TSC is determined only upon encoding the RSL messages. Instead, pick the TSC to use upon the initial lchan activation / modification request; store this in the request structs and pass through the activation / modification code paths. For VAMOS modes, we also need to pick a TSC Set. Do so also upon activ / modif request. Note that the TSC Set is not yet applied in this patch, it will be applied in upcoming VAMOS patches. The activ / modif request may pass -1 for tsc_set and/or tsc to indicate no specific choice of TSC Set and TSC, resulting in the same behavior as before this patch. For example, lchan->activate.info.tsc* may be passed as -1. The exact choice for tsc_set and tsc is then stored in lchan->activate.tsc*, i.e. one level up (the .info sub-struct is considered as immutable input args). The lchan->activate.tsc* are the values actually encoded in RSL messages. After the ACK, the lchan->activate.tsc* is stored in lchan->tsc* to indicate the TSC actually in use. Same for modif. Note that in 3GPP TS 45.002, the TSC Set are numbered 1 to 4, while the TSC are numbered 0 to 7. On the wire, though, TSC Set is sent as 0 to 3 value. This is a weird discrepancy, odd choice made by the spec authors. For conformance with the spec wording, I decided to pass the TSC Set number as a 1-4 ranged value, and only convert it to the 0-3 on-the-wire format upon message encoding. So log messages and VTY output will indicate the first TSC Set as "1", but the first TSC as "0", as defined in 3GPP TS 45.002, even if that is somewhat weird. Related: SYS#5315 OS#4940 Change-Id: Ic665125255d7354f5499d10dda1dd866ab243d24
2021-05-22 14:29:35 +00:00
.tsc_set = req->tsc_set,
.tsc = req->tsc,
add VAMOS secondary lchans to timeslot struct So far there is a bunch of code setting a primary lchan in VAMOS mode. This patch now adds the actual secondary "shadow" lchans that may be combined with a primary lchan in VAMOS mode to form a multiplex. VAMOS lchans are put in the same ts->lchan[] array that keeps the primary lchans. They are at most two additional usable lchans (for a TCH/H shadow) added to either TCH/F or TCH/H. Keeping these in the same array allows looping over all lchans easily. The ts->max_primary_lchans indicates the index of the first VAMOS shadow lchan. Related: SYS#5315 OS#4940 Change-Id: I928af99498bba488d317693f3144d4fccbbe9af3
2021-04-24 12:48:19 +00:00
.vamos = conn->assignment.new_lchan->vamos.is_secondary,
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
};
lchan and assignment FSMs: make Channel Mode Modify more sane The Channel Mode Modify procedure is currently implemented for changing a TCH lchan from signalling to voice mode. For that, however, it is re-using (abusing) the channel activation structs and state transitions, and thus always implies activating a voice stream when the mode modification is done. I will add a Channel Mode Modify to enable VAMOS mode soon, so I require separate structs and state transitions which also work on an lchan that already has a voice stream established: a struct lchan_modify_info and LCHAN_EV_REQUEST_MODE_MODIFY, and dedicated assignment FSM state ASSIGNMENT_ST_WAIT_LCHAN_MODIFIED. For the part where a Channel Mode Modify enables a voice stream after switching from signalling to speech mode, still use the channel activation code path, but only once the mode modification is done. General improvements: - To ask for a mode modification, emit an FSM event that ensures a mode modify only happens when the lchan state allows it. - The new lchan_modify_info struct reflects only those parts that have an effect during a mode modification (before the lchan_activate_info was fully populated, many values not having an effect). - More accurate logging, indicating "Mode Modify" instead of "Channel Activation" A TTCN3 test for the Channel Mode Modify procedure is added in Idf4efaed986de0bbd2b663313e837352cc139f0f, and the test passes both before and after this patch is applied. Related: SYS#4895 Change-Id: I4986844f839b1c9672c61d916eb3d33d0042d747
2021-04-14 17:39:01 +00:00
lchan_activate(conn->assignment.new_lchan, &activ_info);
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
}
assignment_fsm: tweak state transitions (prep for reassignment) It is better design to take state change actions in the onenter function, instead of triggering a state change and then calling lchan_mode_modify() or lchan_activate(). The reason is that in principle, any state change may cause error handling to abort and deallocate FSMs. This is also preparation for reassignment to a specific lchan in an upcoming patch. Related: SYS#5315 OS#4940 OS#3277 Change-Id: I9a2e7eefd4196b80671311e5dfd275893ec0e275
2021-05-18 22:56:07 +00:00
static void assignment_fsm_wait_lchan_active(struct osmo_fsm_inst *fi, uint32_t event, void *data)
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
{
struct gsm_subscriber_connection *conn = assignment_fi_conn(fi);
switch (event) {
case ASSIGNMENT_EV_LCHAN_ACTIVE:
assignment_fsm: tweak state transitions (prep for reassignment) It is better design to take state change actions in the onenter function, instead of triggering a state change and then calling lchan_mode_modify() or lchan_activate(). The reason is that in principle, any state change may cause error handling to abort and deallocate FSMs. This is also preparation for reassignment to a specific lchan in an upcoming patch. Related: SYS#5315 OS#4940 OS#3277 Change-Id: I9a2e7eefd4196b80671311e5dfd275893ec0e275
2021-05-18 22:56:07 +00:00
if (data != conn->assignment.new_lchan) {
LOG_ASSIGNMENT(conn, LOGL_ERROR, "Some unrelated lchan was activated, ignoring: %s\n",
gsm_lchan_name(data));
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
return;
assignment_fsm: tweak state transitions (prep for reassignment) It is better design to take state change actions in the onenter function, instead of triggering a state change and then calling lchan_mode_modify() or lchan_activate(). The reason is that in principle, any state change may cause error handling to abort and deallocate FSMs. This is also preparation for reassignment to a specific lchan in an upcoming patch. Related: SYS#5315 OS#4940 OS#3277 Change-Id: I9a2e7eefd4196b80671311e5dfd275893ec0e275
2021-05-18 22:56:07 +00:00
}
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
/* The TS may have changed its pchan_is */
assignment_fsm_update_id(conn);
assignment_fsm_state_chg(ASSIGNMENT_ST_WAIT_RR_ASS_COMPLETE);
return;
default:
OSMO_ASSERT(false);
}
}
static void assignment_fsm_wait_rr_ass_complete_onenter(struct osmo_fsm_inst *fi, uint32_t prev_state)
{
int rc;
struct gsm_subscriber_connection *conn = assignment_fi_conn(fi);
assignment_fsm: Check for conn->lchan When the SDCCH gets released while the TCH still beeing activated, then the ChanActivACK that is received after the TCH is activated will trigger a segmentation fault in the assignment_fsm. The reason for this is that conn->lchan, which holds the SDCCH at that point in time, is now NULL. To prevent osmo-bsc from crashing, the FSM should check for the presence of conn->lchan first. If it does not exist, the FSM should terminate. (Assignment failed) Change-Id: I3b1cd88bea62ef0032f6c035bac95d3df9fdca7a Related: SYS#5627
2021-09-23 14:19:51 +00:00
/* There may be situations where the SDCCH gets released while the TCH is still being activated. We will then
* receive ChanActivAck message from the BTS when the TCH is ready. Since the SDCCH is already released by
* then conn->lchan will be NULL in this case. */
if (!conn->lchan) {
assignment_fail(GSM0808_CAUSE_EQUIPMENT_FAILURE,
"Unable to send RR Assignment Command: conn without lchan");
return;
}
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
rc = gsm48_send_rr_ass_cmd(conn->lchan, conn->assignment.new_lchan,
conn->lchan->ms_power);
if (rc)
assignment_fail(GSM0808_CAUSE_EQUIPMENT_FAILURE, "Unable to send RR Assignment Command");
}
static uint8_t get_cause(void *data)
{
if (data)
return *(uint8_t*)data;
return GSM0808_CAUSE_EQUIPMENT_FAILURE;
}
static void assignment_fsm_wait_rr_ass_complete(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
struct gsm_subscriber_connection *conn = assignment_fi_conn(fi);
switch (event) {
case ASSIGNMENT_EV_RR_ASSIGNMENT_COMPLETE:
assignment_fsm_state_chg(ASSIGNMENT_ST_WAIT_LCHAN_ESTABLISHED);
return;
case ASSIGNMENT_EV_LCHAN_ESTABLISHED:
cosmetic: FSMs: allow ignorable events In various FSMs, some events may appear later or earlier without need of action. Do not indicate these as 'ERROR' (event not permitted), but allow and ignore them. Debug-log about some of those. From the old code, we've taken over the habit to change into WAIT_BEFORE_RF_RELEASE even before SAPI[0] is released. Hence we may still receive a SAPI[0] REL_IND in WAIT_BEFORE_RF_RELEASE. Don't show this as error message, just silently accept it. Change-Id: Ie320c7c6a1436184aaf2ec5a1843e04f4b3414ab
2018-07-12 03:00:03 +00:00
LOG_ASSIGNMENT(conn, LOGL_DEBUG, "lchan established, still waiting for RR Assignment Complete\n");
/* The lchan is already done with all of its RTP setup. We will notice the lchan state
* being LCHAN_ST_ESTABLISHED in assignment_fsm_wait_lchan_established_onenter(). */
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
return;
case ASSIGNMENT_EV_RR_ASSIGNMENT_FAIL:
Count assignment rates per BTS as well This adds the assignment counters for the BTS as well and changes the assignment_count() macro to increase both the counters for the BSC as well as the BTS. Related: SYS#4877 Change-Id: I0009e51d4caf68e762138d98e2e23d49acc3cc1a
2020-06-08 12:36:46 +00:00
assignment_count_result(CTR_ASSIGNMENT_FAILED);
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
assignment_fail(get_cause(data), "Rx RR Assignment Failure");
return;
default:
OSMO_ASSERT(false);
}
}
static void assignment_fsm_post_lchan_established(struct osmo_fsm_inst *fi);
static void assignment_fsm_wait_lchan_established_onenter(struct osmo_fsm_inst *fi, uint32_t prev_state)
{
struct gsm_subscriber_connection *conn = assignment_fi_conn(fi);
/* Do we still need to wait for the RTP stream at all? */
cosmetic: FSMs: allow ignorable events In various FSMs, some events may appear later or earlier without need of action. Do not indicate these as 'ERROR' (event not permitted), but allow and ignore them. Debug-log about some of those. From the old code, we've taken over the habit to change into WAIT_BEFORE_RF_RELEASE even before SAPI[0] is released. Hence we may still receive a SAPI[0] REL_IND in WAIT_BEFORE_RF_RELEASE. Don't show this as error message, just silently accept it. Change-Id: Ie320c7c6a1436184aaf2ec5a1843e04f4b3414ab
2018-07-12 03:00:03 +00:00
if (lchan_state_is(conn->assignment.new_lchan, LCHAN_ST_ESTABLISHED)) {
fix log line end in assignment_fsm.c Change-Id: I4070ee9164eb161584df70ae174b538c394ab9cd
2019-03-14 22:00:42 +00:00
LOG_ASSIGNMENT(conn, LOGL_DEBUG, "lchan fully established, no need to wait\n");
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
assignment_fsm_post_lchan_established(fi);
cosmetic: FSMs: allow ignorable events In various FSMs, some events may appear later or earlier without need of action. Do not indicate these as 'ERROR' (event not permitted), but allow and ignore them. Debug-log about some of those. From the old code, we've taken over the habit to change into WAIT_BEFORE_RF_RELEASE even before SAPI[0] is released. Hence we may still receive a SAPI[0] REL_IND in WAIT_BEFORE_RF_RELEASE. Don't show this as error message, just silently accept it. Change-Id: Ie320c7c6a1436184aaf2ec5a1843e04f4b3414ab
2018-07-12 03:00:03 +00:00
}
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
}
static void assignment_fsm_wait_lchan_established(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
switch (event) {
case ASSIGNMENT_EV_LCHAN_ESTABLISHED:
assignment_fsm_post_lchan_established(fi);
return;
default:
OSMO_ASSERT(false);
}
}
static void assignment_fsm_post_lchan_established(struct osmo_fsm_inst *fi)
{
struct gsm_subscriber_connection *conn = assignment_fi_conn(fi);
if (conn->assignment.requires_voice_stream)
assignment_fsm_state_chg(ASSIGNMENT_ST_WAIT_MGW_ENDPOINT_TO_MSC);
else
assignment_success(conn);
}
static void assignment_fsm_wait_mgw_endpoint_to_msc_onenter(struct osmo_fsm_inst *fi, uint32_t prev_state)
{
struct gsm_subscriber_connection *conn = assignment_fi_conn(fi);
OSMO_ASSERT(conn->assignment.requires_voice_stream);
LOG_ASSIGNMENT(conn, LOGL_DEBUG,
"Connecting MGW endpoint to the MSC's RTP port: %s:%u\n",
conn->assignment.req.msc_rtp_addr,
conn->assignment.req.msc_rtp_port);
lchan and assignment FSMs: make Channel Mode Modify more sane The Channel Mode Modify procedure is currently implemented for changing a TCH lchan from signalling to voice mode. For that, however, it is re-using (abusing) the channel activation structs and state transitions, and thus always implies activating a voice stream when the mode modification is done. I will add a Channel Mode Modify to enable VAMOS mode soon, so I require separate structs and state transitions which also work on an lchan that already has a voice stream established: a struct lchan_modify_info and LCHAN_EV_REQUEST_MODE_MODIFY, and dedicated assignment FSM state ASSIGNMENT_ST_WAIT_LCHAN_MODIFIED. For the part where a Channel Mode Modify enables a voice stream after switching from signalling to speech mode, still use the channel activation code path, but only once the mode modification is done. General improvements: - To ask for a mode modification, emit an FSM event that ensures a mode modify only happens when the lchan state allows it. - The new lchan_modify_info struct reflects only those parts that have an effect during a mode modification (before the lchan_activate_info was fully populated, many values not having an effect). - More accurate logging, indicating "Mode Modify" instead of "Channel Activation" A TTCN3 test for the Channel Mode Modify procedure is added in Idf4efaed986de0bbd2b663313e837352cc139f0f, and the test passes both before and after this patch is applied. Related: SYS#4895 Change-Id: I4986844f839b1c9672c61d916eb3d33d0042d747
2021-04-14 17:39:01 +00:00
/* Assignment can do a new channel activation, in which case new_lchan points at the new lchan.
* Or assignment can Channel Mode Modify the already used lchan, in which case new_lchan == NULL. */
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
if (!gscon_connect_mgw_to_msc(conn,
lchan and assignment FSMs: make Channel Mode Modify more sane The Channel Mode Modify procedure is currently implemented for changing a TCH lchan from signalling to voice mode. For that, however, it is re-using (abusing) the channel activation structs and state transitions, and thus always implies activating a voice stream when the mode modification is done. I will add a Channel Mode Modify to enable VAMOS mode soon, so I require separate structs and state transitions which also work on an lchan that already has a voice stream established: a struct lchan_modify_info and LCHAN_EV_REQUEST_MODE_MODIFY, and dedicated assignment FSM state ASSIGNMENT_ST_WAIT_LCHAN_MODIFIED. For the part where a Channel Mode Modify enables a voice stream after switching from signalling to speech mode, still use the channel activation code path, but only once the mode modification is done. General improvements: - To ask for a mode modification, emit an FSM event that ensures a mode modify only happens when the lchan state allows it. - The new lchan_modify_info struct reflects only those parts that have an effect during a mode modification (before the lchan_activate_info was fully populated, many values not having an effect). - More accurate logging, indicating "Mode Modify" instead of "Channel Activation" A TTCN3 test for the Channel Mode Modify procedure is added in Idf4efaed986de0bbd2b663313e837352cc139f0f, and the test passes both before and after this patch is applied. Related: SYS#4895 Change-Id: I4986844f839b1c9672c61d916eb3d33d0042d747
2021-04-14 17:39:01 +00:00
conn->assignment.new_lchan ? : conn->lchan,
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
conn->assignment.req.msc_rtp_addr,
conn->assignment.req.msc_rtp_port,
fi,
ASSIGNMENT_EV_MSC_MGW_OK,
ASSIGNMENT_EV_MSC_MGW_FAIL,
NULL,
&conn->assignment.created_ci_for_msc)) {
assignment_fail(GSM0808_CAUSE_EQUIPMENT_FAILURE,
"Unable to connect MGW endpoint to the MSC side");
return;
}
}
static void assignment_fsm_wait_mgw_endpoint_to_msc(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
struct gsm_subscriber_connection *conn = assignment_fi_conn(fi);
switch (event) {
case ASSIGNMENT_EV_MSC_MGW_OK:
/* For AoIP, we created the MGW endpoint. Ensure it is really there, and log it. */
if (gscon_is_aoip(conn)) {
const struct mgcp_conn_peer *mgw_info;
move mgw endpoint FSM to osmo-mgw.git osmo-mgw.git also includes fixes of the MGW endpoint FSM, for example I92a9944acc96398acd6649f9c3c5badec5dd6dcc. Depends: I9a3effd38e72841529df6c135c077116981dea36 (osmo-mgw) Change-Id: I03e6b48d9b0a5370310d5f56809259ff7909cf9d
2019-04-23 16:37:37 +00:00
mgw_info = osmo_mgcpc_ep_ci_get_rtp_info(conn->user_plane.mgw_endpoint_ci_msc);
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
if (!mgw_info) {
assignment_fail(GSM0808_CAUSE_EQUIPMENT_FAILURE,
"Unable to retrieve RTP port info allocated by MGW for"
" the MSC side.");
return;
}
LOG_ASSIGNMENT(conn, LOGL_DEBUG, "MGW's MSC side CI: %s:%u\n",
mgw_info->addr, mgw_info->port);
}
assignment_success(conn);
return;
case ASSIGNMENT_EV_MSC_MGW_FAIL:
assignment_fail(GSM0808_CAUSE_EQUIPMENT_FAILURE,
"Unable to connect MGW endpoint to the MSC side");
return;
default:
OSMO_ASSERT(false);
}
}
assignment_fsm: tweak state transitions (prep for reassignment) It is better design to take state change actions in the onenter function, instead of triggering a state change and then calling lchan_mode_modify() or lchan_activate(). The reason is that in principle, any state change may cause error handling to abort and deallocate FSMs. This is also preparation for reassignment to a specific lchan in an upcoming patch. Related: SYS#5315 OS#4940 OS#3277 Change-Id: I9a2e7eefd4196b80671311e5dfd275893ec0e275
2021-05-18 22:56:07 +00:00
static void assignment_fsm_wait_lchan_modified_onenter(struct osmo_fsm_inst *fi, uint32_t prev_state)
{
struct gsm_subscriber_connection *conn = assignment_fi_conn(fi);
fixup: pass tsc = -1 for previous default training sequence code An unintended change in default behavior was introduced in patch: "allow explixit TSC Set and TSC on chan activ / modif / assignment" Ic665125255d7354f5499d10dda1dd866ab243d24 c33eb8d56943b8981523754b081967e6ff5f245d Set tsc_set and tsc = -1 for all lchan_activate_info and assignment_request requests to actually yield the default behavior of selecting the TSC based on the timeslot cfg or the BSIC value. By setting tsc = 0 implicitly, the patch caused all requests to ask for tsc 0 instead of calling gsm_ts_tsc(). For a Channel Mode Modify in assignment_fsm, pass the lchan's current TSC to keep it unchanged. osmo-ttcn3-hacks Id67a949e0f61ec8123976eb8d336f04510c55c01 adds a test to verify the expected TSC in all of the activation, assignment and modify messages. Current osmo-bsc master fails, this patch fixes. Related: SYS#5315 OS#4940 Ic665125255d7354f5499d10dda1dd866ab243d24 Change-Id: If12df11511fe22ea167782f776736a1a9c484b1f
2021-06-04 15:53:37 +00:00
struct gsm_lchan *lchan = conn->lchan;
assignment_fsm: tweak state transitions (prep for reassignment) It is better design to take state change actions in the onenter function, instead of triggering a state change and then calling lchan_mode_modify() or lchan_activate(). The reason is that in principle, any state change may cause error handling to abort and deallocate FSMs. This is also preparation for reassignment to a specific lchan in an upcoming patch. Related: SYS#5315 OS#4940 OS#3277 Change-Id: I9a2e7eefd4196b80671311e5dfd275893ec0e275
2021-05-18 22:56:07 +00:00
struct assignment_request *req = &conn->assignment.req;
struct lchan_modify_info modif_info = {
.modify_for = MODIFY_FOR_ASSIGNMENT,
.ch_mode_rate = conn->assignment.selected_ch_mode_rate,
.requires_voice_stream = conn->assignment.requires_voice_stream,
.msc_assigned_cic = req->msc_assigned_cic,
tsc fixup: tweak condition for coverity Related: CID#240352 SYS#4895 Change-Id: Idbc6b47df0aeca9b6d870386b037dd6fb7c96c82
2021-10-05 06:28:03 +00:00
/* keep previous training sequence code. TSC is always present, TSC Set may or may not be an explicit
* value. */
fix TSC / TSC Set used for Handover From the nature of the lchan_activate_info.tsc_set and .tsc, it is easy to forget to set tsc_set,tsc = -1 to use default TSC Set and TSC values. Handover code is one such instance that forgets to set -1. Change the semantics of tsc_set and tsc so that this kind of error can not happen again as easily: use a separate bool to flag whether to use the default config or explicit values. Implicitly fix the lchan_activate_infos "launched" in handover_fsm.c as well as abis_rsl_chan_rqd_queue_poll(). Related: OS#5244 SYS#4895 Related: I1ed6f068c85b01e5a2d7b5f2651498a1521f89af (osmo-ttcn3-hacks) Change-Id: Iae20df4387c3d75752301bd5daeeea7508966393
2021-10-02 10:03:19 +00:00
.tsc_set = {
.present = (lchan->tsc_set >= 0),
.val = lchan->tsc_set,
},
.tsc = {
tsc fixup: tweak condition for coverity Related: CID#240352 SYS#4895 Change-Id: Idbc6b47df0aeca9b6d870386b037dd6fb7c96c82
2021-10-05 06:28:03 +00:00
.present = true,
fix TSC / TSC Set used for Handover From the nature of the lchan_activate_info.tsc_set and .tsc, it is easy to forget to set tsc_set,tsc = -1 to use default TSC Set and TSC values. Handover code is one such instance that forgets to set -1. Change the semantics of tsc_set and tsc so that this kind of error can not happen again as easily: use a separate bool to flag whether to use the default config or explicit values. Implicitly fix the lchan_activate_infos "launched" in handover_fsm.c as well as abis_rsl_chan_rqd_queue_poll(). Related: OS#5244 SYS#4895 Related: I1ed6f068c85b01e5a2d7b5f2651498a1521f89af (osmo-ttcn3-hacks) Change-Id: Iae20df4387c3d75752301bd5daeeea7508966393
2021-10-02 10:03:19 +00:00
.val = lchan->tsc,
},
assignment_fsm: tweak state transitions (prep for reassignment) It is better design to take state change actions in the onenter function, instead of triggering a state change and then calling lchan_mode_modify() or lchan_activate(). The reason is that in principle, any state change may cause error handling to abort and deallocate FSMs. This is also preparation for reassignment to a specific lchan in an upcoming patch. Related: SYS#5315 OS#4940 OS#3277 Change-Id: I9a2e7eefd4196b80671311e5dfd275893ec0e275
2021-05-18 22:56:07 +00:00
};
fixup: pass tsc = -1 for previous default training sequence code An unintended change in default behavior was introduced in patch: "allow explixit TSC Set and TSC on chan activ / modif / assignment" Ic665125255d7354f5499d10dda1dd866ab243d24 c33eb8d56943b8981523754b081967e6ff5f245d Set tsc_set and tsc = -1 for all lchan_activate_info and assignment_request requests to actually yield the default behavior of selecting the TSC based on the timeslot cfg or the BSIC value. By setting tsc = 0 implicitly, the patch caused all requests to ask for tsc 0 instead of calling gsm_ts_tsc(). For a Channel Mode Modify in assignment_fsm, pass the lchan's current TSC to keep it unchanged. osmo-ttcn3-hacks Id67a949e0f61ec8123976eb8d336f04510c55c01 adds a test to verify the expected TSC in all of the activation, assignment and modify messages. Current osmo-bsc master fails, this patch fixes. Related: SYS#5315 OS#4940 Ic665125255d7354f5499d10dda1dd866ab243d24 Change-Id: If12df11511fe22ea167782f776736a1a9c484b1f
2021-06-04 15:53:37 +00:00
lchan_mode_modify(lchan, &modif_info);
assignment_fsm: tweak state transitions (prep for reassignment) It is better design to take state change actions in the onenter function, instead of triggering a state change and then calling lchan_mode_modify() or lchan_activate(). The reason is that in principle, any state change may cause error handling to abort and deallocate FSMs. This is also preparation for reassignment to a specific lchan in an upcoming patch. Related: SYS#5315 OS#4940 OS#3277 Change-Id: I9a2e7eefd4196b80671311e5dfd275893ec0e275
2021-05-18 22:56:07 +00:00
}
lchan and assignment FSMs: make Channel Mode Modify more sane The Channel Mode Modify procedure is currently implemented for changing a TCH lchan from signalling to voice mode. For that, however, it is re-using (abusing) the channel activation structs and state transitions, and thus always implies activating a voice stream when the mode modification is done. I will add a Channel Mode Modify to enable VAMOS mode soon, so I require separate structs and state transitions which also work on an lchan that already has a voice stream established: a struct lchan_modify_info and LCHAN_EV_REQUEST_MODE_MODIFY, and dedicated assignment FSM state ASSIGNMENT_ST_WAIT_LCHAN_MODIFIED. For the part where a Channel Mode Modify enables a voice stream after switching from signalling to speech mode, still use the channel activation code path, but only once the mode modification is done. General improvements: - To ask for a mode modification, emit an FSM event that ensures a mode modify only happens when the lchan state allows it. - The new lchan_modify_info struct reflects only those parts that have an effect during a mode modification (before the lchan_activate_info was fully populated, many values not having an effect). - More accurate logging, indicating "Mode Modify" instead of "Channel Activation" A TTCN3 test for the Channel Mode Modify procedure is added in Idf4efaed986de0bbd2b663313e837352cc139f0f, and the test passes both before and after this patch is applied. Related: SYS#4895 Change-Id: I4986844f839b1c9672c61d916eb3d33d0042d747
2021-04-14 17:39:01 +00:00
static void assignment_fsm_wait_lchan_modified(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
switch (event) {
case ASSIGNMENT_EV_LCHAN_MODIFIED:
assignment_fsm_post_lchan_established(fi);
return;
default:
OSMO_ASSERT(false);
}
}
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
#define S(x) (1 << (x))
static const struct osmo_fsm_state assignment_fsm_states[] = {
[ASSIGNMENT_ST_WAIT_LCHAN_ACTIVE] = {
.name = "WAIT_LCHAN_ACTIVE",
assignment_fsm: tweak state transitions (prep for reassignment) It is better design to take state change actions in the onenter function, instead of triggering a state change and then calling lchan_mode_modify() or lchan_activate(). The reason is that in principle, any state change may cause error handling to abort and deallocate FSMs. This is also preparation for reassignment to a specific lchan in an upcoming patch. Related: SYS#5315 OS#4940 OS#3277 Change-Id: I9a2e7eefd4196b80671311e5dfd275893ec0e275
2021-05-18 22:56:07 +00:00
.onenter = assignment_fsm_wait_lchan_active_onenter,
.action = assignment_fsm_wait_lchan_active,
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
.in_event_mask = 0
| S(ASSIGNMENT_EV_LCHAN_ACTIVE)
,
.out_state_mask = 0
| S(ASSIGNMENT_ST_WAIT_LCHAN_ACTIVE)
| S(ASSIGNMENT_ST_WAIT_RR_ASS_COMPLETE)
lchan and assignment FSMs: make Channel Mode Modify more sane The Channel Mode Modify procedure is currently implemented for changing a TCH lchan from signalling to voice mode. For that, however, it is re-using (abusing) the channel activation structs and state transitions, and thus always implies activating a voice stream when the mode modification is done. I will add a Channel Mode Modify to enable VAMOS mode soon, so I require separate structs and state transitions which also work on an lchan that already has a voice stream established: a struct lchan_modify_info and LCHAN_EV_REQUEST_MODE_MODIFY, and dedicated assignment FSM state ASSIGNMENT_ST_WAIT_LCHAN_MODIFIED. For the part where a Channel Mode Modify enables a voice stream after switching from signalling to speech mode, still use the channel activation code path, but only once the mode modification is done. General improvements: - To ask for a mode modification, emit an FSM event that ensures a mode modify only happens when the lchan state allows it. - The new lchan_modify_info struct reflects only those parts that have an effect during a mode modification (before the lchan_activate_info was fully populated, many values not having an effect). - More accurate logging, indicating "Mode Modify" instead of "Channel Activation" A TTCN3 test for the Channel Mode Modify procedure is added in Idf4efaed986de0bbd2b663313e837352cc139f0f, and the test passes both before and after this patch is applied. Related: SYS#4895 Change-Id: I4986844f839b1c9672c61d916eb3d33d0042d747
2021-04-14 17:39:01 +00:00
| S(ASSIGNMENT_ST_WAIT_LCHAN_MODIFIED)
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
,
},
[ASSIGNMENT_ST_WAIT_RR_ASS_COMPLETE] = {
.name = "WAIT_RR_ASS_COMPLETE",
.onenter = assignment_fsm_wait_rr_ass_complete_onenter,
.action = assignment_fsm_wait_rr_ass_complete,
.in_event_mask = 0
| S(ASSIGNMENT_EV_RR_ASSIGNMENT_COMPLETE)
| S(ASSIGNMENT_EV_RR_ASSIGNMENT_FAIL)
cosmetic: FSMs: allow ignorable events In various FSMs, some events may appear later or earlier without need of action. Do not indicate these as 'ERROR' (event not permitted), but allow and ignore them. Debug-log about some of those. From the old code, we've taken over the habit to change into WAIT_BEFORE_RF_RELEASE even before SAPI[0] is released. Hence we may still receive a SAPI[0] REL_IND in WAIT_BEFORE_RF_RELEASE. Don't show this as error message, just silently accept it. Change-Id: Ie320c7c6a1436184aaf2ec5a1843e04f4b3414ab
2018-07-12 03:00:03 +00:00
| S(ASSIGNMENT_EV_LCHAN_ESTABLISHED)
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
,
.out_state_mask = 0
| S(ASSIGNMENT_ST_WAIT_LCHAN_ESTABLISHED)
,
},
[ASSIGNMENT_ST_WAIT_LCHAN_ESTABLISHED] = {
.name = "WAIT_LCHAN_ESTABLISHED",
.onenter = assignment_fsm_wait_lchan_established_onenter,
.action = assignment_fsm_wait_lchan_established,
.in_event_mask = 0
| S(ASSIGNMENT_EV_LCHAN_ESTABLISHED)
,
.out_state_mask = 0
| S(ASSIGNMENT_ST_WAIT_MGW_ENDPOINT_TO_MSC)
,
},
[ASSIGNMENT_ST_WAIT_MGW_ENDPOINT_TO_MSC] = {
.name = "WAIT_MGW_ENDPOINT_TO_MSC",
.onenter = assignment_fsm_wait_mgw_endpoint_to_msc_onenter,
.action = assignment_fsm_wait_mgw_endpoint_to_msc,
.in_event_mask = 0
| S(ASSIGNMENT_EV_MSC_MGW_OK)
| S(ASSIGNMENT_EV_MSC_MGW_FAIL)
,
},
lchan and assignment FSMs: make Channel Mode Modify more sane The Channel Mode Modify procedure is currently implemented for changing a TCH lchan from signalling to voice mode. For that, however, it is re-using (abusing) the channel activation structs and state transitions, and thus always implies activating a voice stream when the mode modification is done. I will add a Channel Mode Modify to enable VAMOS mode soon, so I require separate structs and state transitions which also work on an lchan that already has a voice stream established: a struct lchan_modify_info and LCHAN_EV_REQUEST_MODE_MODIFY, and dedicated assignment FSM state ASSIGNMENT_ST_WAIT_LCHAN_MODIFIED. For the part where a Channel Mode Modify enables a voice stream after switching from signalling to speech mode, still use the channel activation code path, but only once the mode modification is done. General improvements: - To ask for a mode modification, emit an FSM event that ensures a mode modify only happens when the lchan state allows it. - The new lchan_modify_info struct reflects only those parts that have an effect during a mode modification (before the lchan_activate_info was fully populated, many values not having an effect). - More accurate logging, indicating "Mode Modify" instead of "Channel Activation" A TTCN3 test for the Channel Mode Modify procedure is added in Idf4efaed986de0bbd2b663313e837352cc139f0f, and the test passes both before and after this patch is applied. Related: SYS#4895 Change-Id: I4986844f839b1c9672c61d916eb3d33d0042d747
2021-04-14 17:39:01 +00:00
[ASSIGNMENT_ST_WAIT_LCHAN_MODIFIED] = {
.name = "WAIT_LCHAN_MODIFIED",
assignment_fsm: tweak state transitions (prep for reassignment) It is better design to take state change actions in the onenter function, instead of triggering a state change and then calling lchan_mode_modify() or lchan_activate(). The reason is that in principle, any state change may cause error handling to abort and deallocate FSMs. This is also preparation for reassignment to a specific lchan in an upcoming patch. Related: SYS#5315 OS#4940 OS#3277 Change-Id: I9a2e7eefd4196b80671311e5dfd275893ec0e275
2021-05-18 22:56:07 +00:00
.onenter = assignment_fsm_wait_lchan_modified_onenter,
lchan and assignment FSMs: make Channel Mode Modify more sane The Channel Mode Modify procedure is currently implemented for changing a TCH lchan from signalling to voice mode. For that, however, it is re-using (abusing) the channel activation structs and state transitions, and thus always implies activating a voice stream when the mode modification is done. I will add a Channel Mode Modify to enable VAMOS mode soon, so I require separate structs and state transitions which also work on an lchan that already has a voice stream established: a struct lchan_modify_info and LCHAN_EV_REQUEST_MODE_MODIFY, and dedicated assignment FSM state ASSIGNMENT_ST_WAIT_LCHAN_MODIFIED. For the part where a Channel Mode Modify enables a voice stream after switching from signalling to speech mode, still use the channel activation code path, but only once the mode modification is done. General improvements: - To ask for a mode modification, emit an FSM event that ensures a mode modify only happens when the lchan state allows it. - The new lchan_modify_info struct reflects only those parts that have an effect during a mode modification (before the lchan_activate_info was fully populated, many values not having an effect). - More accurate logging, indicating "Mode Modify" instead of "Channel Activation" A TTCN3 test for the Channel Mode Modify procedure is added in Idf4efaed986de0bbd2b663313e837352cc139f0f, and the test passes both before and after this patch is applied. Related: SYS#4895 Change-Id: I4986844f839b1c9672c61d916eb3d33d0042d747
2021-04-14 17:39:01 +00:00
.action = assignment_fsm_wait_lchan_modified,
.in_event_mask = 0
| S(ASSIGNMENT_EV_LCHAN_MODIFIED)
,
.out_state_mask = 0
| S(ASSIGNMENT_ST_WAIT_MGW_ENDPOINT_TO_MSC)
,
},
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
};
static const struct value_string assignment_fsm_event_names[] = {
OSMO_VALUE_STRING(ASSIGNMENT_EV_LCHAN_ACTIVE),
OSMO_VALUE_STRING(ASSIGNMENT_EV_LCHAN_ESTABLISHED),
lchan and assignment FSMs: make Channel Mode Modify more sane The Channel Mode Modify procedure is currently implemented for changing a TCH lchan from signalling to voice mode. For that, however, it is re-using (abusing) the channel activation structs and state transitions, and thus always implies activating a voice stream when the mode modification is done. I will add a Channel Mode Modify to enable VAMOS mode soon, so I require separate structs and state transitions which also work on an lchan that already has a voice stream established: a struct lchan_modify_info and LCHAN_EV_REQUEST_MODE_MODIFY, and dedicated assignment FSM state ASSIGNMENT_ST_WAIT_LCHAN_MODIFIED. For the part where a Channel Mode Modify enables a voice stream after switching from signalling to speech mode, still use the channel activation code path, but only once the mode modification is done. General improvements: - To ask for a mode modification, emit an FSM event that ensures a mode modify only happens when the lchan state allows it. - The new lchan_modify_info struct reflects only those parts that have an effect during a mode modification (before the lchan_activate_info was fully populated, many values not having an effect). - More accurate logging, indicating "Mode Modify" instead of "Channel Activation" A TTCN3 test for the Channel Mode Modify procedure is added in Idf4efaed986de0bbd2b663313e837352cc139f0f, and the test passes both before and after this patch is applied. Related: SYS#4895 Change-Id: I4986844f839b1c9672c61d916eb3d33d0042d747
2021-04-14 17:39:01 +00:00
OSMO_VALUE_STRING(ASSIGNMENT_EV_LCHAN_MODIFIED),
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
OSMO_VALUE_STRING(ASSIGNMENT_EV_LCHAN_ERROR),
OSMO_VALUE_STRING(ASSIGNMENT_EV_MSC_MGW_OK),
OSMO_VALUE_STRING(ASSIGNMENT_EV_MSC_MGW_FAIL),
OSMO_VALUE_STRING(ASSIGNMENT_EV_RR_ASSIGNMENT_COMPLETE),
OSMO_VALUE_STRING(ASSIGNMENT_EV_RR_ASSIGNMENT_FAIL),
OSMO_VALUE_STRING(ASSIGNMENT_EV_CONN_RELEASING),
{}
};
assignment_fsm: make assignment_fsm_allstate_action static assignment_fsm_allstate_action is only used in module assignment_fsm.c, so lets decalare it as static Change-Id: I90d6c11bd485e3be80efc7f146e0129d93e18458
2021-09-27 10:45:23 +00:00
static void assignment_fsm_allstate_action(struct osmo_fsm_inst *fi, uint32_t event, void *data)
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
{
struct gsm_subscriber_connection *conn = assignment_fi_conn(fi);
lchan and assignment FSMs: make Channel Mode Modify more sane The Channel Mode Modify procedure is currently implemented for changing a TCH lchan from signalling to voice mode. For that, however, it is re-using (abusing) the channel activation structs and state transitions, and thus always implies activating a voice stream when the mode modification is done. I will add a Channel Mode Modify to enable VAMOS mode soon, so I require separate structs and state transitions which also work on an lchan that already has a voice stream established: a struct lchan_modify_info and LCHAN_EV_REQUEST_MODE_MODIFY, and dedicated assignment FSM state ASSIGNMENT_ST_WAIT_LCHAN_MODIFIED. For the part where a Channel Mode Modify enables a voice stream after switching from signalling to speech mode, still use the channel activation code path, but only once the mode modification is done. General improvements: - To ask for a mode modification, emit an FSM event that ensures a mode modify only happens when the lchan state allows it. - The new lchan_modify_info struct reflects only those parts that have an effect during a mode modification (before the lchan_activate_info was fully populated, many values not having an effect). - More accurate logging, indicating "Mode Modify" instead of "Channel Activation" A TTCN3 test for the Channel Mode Modify procedure is added in Idf4efaed986de0bbd2b663313e837352cc139f0f, and the test passes both before and after this patch is applied. Related: SYS#4895 Change-Id: I4986844f839b1c9672c61d916eb3d33d0042d747
2021-04-14 17:39:01 +00:00
/* Assignment can do a new channel activation, in which case new_lchan points at the new lchan.
* Or assignment can Channel Mode Modify the already used lchan, in which case new_lchan == NULL. */
struct gsm_lchan *new_lchan = conn->assignment.new_lchan ? : conn->lchan;
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
switch (event) {
case ASSIGNMENT_EV_CONN_RELEASING:
Count assignment rates per BTS as well This adds the assignment counters for the BTS as well and changes the assignment_count() macro to increase both the counters for the BSC as well as the BTS. Related: SYS#4877 Change-Id: I0009e51d4caf68e762138d98e2e23d49acc3cc1a
2020-06-08 12:36:46 +00:00
assignment_count_result(CTR_ASSIGNMENT_STOPPED);
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
osmo_fsm_inst_term(fi, OSMO_FSM_TERM_REQUEST, 0);
return;
case ASSIGNMENT_EV_LCHAN_ERROR:
lchan and assignment FSMs: make Channel Mode Modify more sane The Channel Mode Modify procedure is currently implemented for changing a TCH lchan from signalling to voice mode. For that, however, it is re-using (abusing) the channel activation structs and state transitions, and thus always implies activating a voice stream when the mode modification is done. I will add a Channel Mode Modify to enable VAMOS mode soon, so I require separate structs and state transitions which also work on an lchan that already has a voice stream established: a struct lchan_modify_info and LCHAN_EV_REQUEST_MODE_MODIFY, and dedicated assignment FSM state ASSIGNMENT_ST_WAIT_LCHAN_MODIFIED. For the part where a Channel Mode Modify enables a voice stream after switching from signalling to speech mode, still use the channel activation code path, but only once the mode modification is done. General improvements: - To ask for a mode modification, emit an FSM event that ensures a mode modify only happens when the lchan state allows it. - The new lchan_modify_info struct reflects only those parts that have an effect during a mode modification (before the lchan_activate_info was fully populated, many values not having an effect). - More accurate logging, indicating "Mode Modify" instead of "Channel Activation" A TTCN3 test for the Channel Mode Modify procedure is added in Idf4efaed986de0bbd2b663313e837352cc139f0f, and the test passes both before and after this patch is applied. Related: SYS#4895 Change-Id: I4986844f839b1c9672c61d916eb3d33d0042d747
2021-04-14 17:39:01 +00:00
if (data != new_lchan)
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
return;
lchan and assignment FSMs: make Channel Mode Modify more sane The Channel Mode Modify procedure is currently implemented for changing a TCH lchan from signalling to voice mode. For that, however, it is re-using (abusing) the channel activation structs and state transitions, and thus always implies activating a voice stream when the mode modification is done. I will add a Channel Mode Modify to enable VAMOS mode soon, so I require separate structs and state transitions which also work on an lchan that already has a voice stream established: a struct lchan_modify_info and LCHAN_EV_REQUEST_MODE_MODIFY, and dedicated assignment FSM state ASSIGNMENT_ST_WAIT_LCHAN_MODIFIED. For the part where a Channel Mode Modify enables a voice stream after switching from signalling to speech mode, still use the channel activation code path, but only once the mode modification is done. General improvements: - To ask for a mode modification, emit an FSM event that ensures a mode modify only happens when the lchan state allows it. - The new lchan_modify_info struct reflects only those parts that have an effect during a mode modification (before the lchan_activate_info was fully populated, many values not having an effect). - More accurate logging, indicating "Mode Modify" instead of "Channel Activation" A TTCN3 test for the Channel Mode Modify procedure is added in Idf4efaed986de0bbd2b663313e837352cc139f0f, and the test passes both before and after this patch is applied. Related: SYS#4895 Change-Id: I4986844f839b1c9672c61d916eb3d33d0042d747
2021-04-14 17:39:01 +00:00
assignment_fail(new_lchan->activate.gsm0808_error_cause,
"Failed to %s lchan %s",
conn->assignment.new_lchan ? "activate" : "modify",
assignment_fsm: Log modified lchan in assignment_fsm_allstate_action() Change-Id: I579c265e9d18402ce806e4d0ea836042185dedaa
2021-11-04 21:09:41 +00:00
gsm_lchan_name(new_lchan));
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
return;
default:
return;
}
}
assignment_fsm: make assignment_fsm_timer_cb static assignment_fsm_timer_cb is only used in module assignment_fsm.c, so lets decalare it as static Change-Id: Ibb3aa01d21b0a83600dc22382a5c8001debbe054
2021-09-27 10:43:59 +00:00
static int assignment_fsm_timer_cb(struct osmo_fsm_inst *fi)
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
{
struct gsm_subscriber_connection *conn = assignment_fi_conn(fi);
Count assignment rates per BTS as well This adds the assignment counters for the BTS as well and changes the assignment_count() macro to increase both the counters for the BSC as well as the BTS. Related: SYS#4877 Change-Id: I0009e51d4caf68e762138d98e2e23d49acc3cc1a
2020-06-08 12:36:46 +00:00
assignment_count_result(CTR_ASSIGNMENT_TIMEOUT);
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
assignment_fail(GSM0808_CAUSE_EQUIPMENT_FAILURE, "Timeout");
return 0;
}
void assignment_fsm_cleanup(struct osmo_fsm_inst *fi, enum osmo_fsm_term_cause cause)
{
struct gsm_subscriber_connection *conn = assignment_fi_conn(fi);
assignment_reset(conn);
conn->assignment.fi = NULL;
}
static struct osmo_fsm assignment_fsm = {
.name = "assignment",
.states = assignment_fsm_states,
.num_states = ARRAY_SIZE(assignment_fsm_states),
create separate logging categories for lchan,ts,as FSMs Change-Id: Ie889b8860a4a63c7c22ef65025f690d64cd7330c
2018-07-11 17:53:39 +00:00
.log_subsys = DAS,
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
.event_names = assignment_fsm_event_names,
.allstate_action = assignment_fsm_allstate_action,
.allstate_event_mask = 0
| S(ASSIGNMENT_EV_CONN_RELEASING)
| S(ASSIGNMENT_EV_LCHAN_ERROR)
,
.timer_cb = assignment_fsm_timer_cb,
.cleanup = assignment_fsm_cleanup,
};